drought flood case study

Publication: Climate Change Impacts in Drought and Flood Affected Areas : Case Studies in India

Files in English

Link to data set, report series, other publications in this report series, journal volume, journal issue, collections, associated urls, associated content.

Kenya’s devastating floods expose decades of poor urban planning and bad land management

Chartered Consultant in Hydrology and Water Resources, Visiting Research Fellow, King's College London

Disclosure statement

Sean Avery is affiliated with: Hydrological Society of Kenya, Water Resource Associates, Kings College London, University of Gent

King's College London provides funding as a member of The Conversation UK.

View all partners

Floods in Kenya killed at least 169 people between March and April 2024. The most catastrophic of these deaths occurred after a flash flood swept through a rural village killing 42 people . Death and destruction have also occurred in the capital, Nairobi, a stark reminder of the persistent failure to keep abreast of the city’s rapid urbanisation needs. Sean Avery , who has undertaken numerous flood and drainage studies throughout Africa, unpacks the problems and potential solutions.

Are floods in Kenya causing more damage? If so, why?

Floods are the natural consequence of storm rainfall and have an important ecological role . They inundate flood plains where silts settle, riverbed aquifers are recharged and nutrients are gathered. Annual rainfall in Kenya varies from 2,000mm in the western region to less than 250mm in the drylands covering over 80% of Kenya. But storm rainfalls are widespread. This means that floods can occur in any part of the country.

The impact of floods has become more severe due to a number of factors.

The first is how much water runs off. In rural areas, changes to the landscape have meant that there’s been an increase in the amount of storm runoff generated from rainfall. This is because the natural state of the land has been altered through settlement, roads, deforestation, livestock grazing and cultivation. As a result, a greater proportion of rainfall runs off. This runoff is more rapid and erosive, and less water infiltrates to replenish groundwater stores.

The East African Flood Model , a standard drainage design tool, demonstrates that by reducing a forested catchment into a field for livestock pasture, for instance, the peak flood magnitude can increase 20-fold. This form of catchment degradation leads to landslides, dams can breach, and road culverts and irrigation intakes are regularly washed away.

Land degradation in sub-Saharan rangelands is omnipresent, with over 90% rangeland degradation reported in Kenya’s northern drylands . Kenyan research has recorded dramatic increases in stormwater runoff due to overgrazing.

Second, human pressure in urban areas – including encroachment into riparian zones and loss of natural flood storage buffers through the destruction of wetlands – has increased flood risks. Riparian zones are areas bordering rivers and other bodies of water.

By 2050, half of Kenya’s population will live in urban areas. Green space is progressively being filled with buildings and pavements. A large proportion of urban population lives in tin-roofed slums and informal settlements lacking adequate drainage infrastructure. As a result, almost all of the storm rainfall is translated into rapid and sometimes catastrophic flooding.

Third, flood risks are worse for people who have settled in vacant land which is often in low-lying areas and within flood plains. In these areas, inundation by flood waters is inevitable.

Fourth, Nairobi’s persistent water supply shortages have led to a proliferation of boreholes whose over-abstraction has resulted in a dramatic decline in the underground water table’s levels. This leads to aquifer compression, which is compounded by the weight of buildings. The result is ground level subsidence , which creates low spots where stormwater floods collect.

What should be done to minimise the risks?

Rural areas require a different set of solutions.

Natural watercourses throughout Kenya are being scoured out by larger floods due to land use pressures. These watercourses are expanding and riparian vegetation cover is disappearing. The flood plains need space to regenerate the natural vegetation cover as this attenuates floods, reducing the force of runoff and erosion.

There are existing laws to protect riverbanks, and livestock movements in these areas must also be controlled. Any building or informal settlement within riparian areas is illegal and would otherwise be exposed to the dangers of floods. Enforcement is a challenge, however, as these areas are favoured by human activities and often these people are among the poorest.

Urban areas have a host of particular challenges that need to be addressed.

Take Nairobi, Kenya’s capital city. The physical planning process is hindered by corruption . Inappropriate and unsafe developments proliferate alongside inadequate water supply, wastewater and solid waste disposal infrastructure. Sewage effluent is often discharged into stormwater drains, even in high-class areas of the city. And there is little control of development in the growing urban centres bordering Nairobi, with transport corridors being congested. Throughout the country, laws that protect riparian zones are flouted.

None of this is sustainable.

Each municipality is obliged to provide infrastructure that includes an effective engineered stormwater drainage network. And in parallel, wastewater and solid wastes must be separately managed.

The typical stormwater drainage network comprises adequately sized earth and lined channels, and pipes and culverts that convey the stormwater to the nearest watercourse. Constant maintenance is essential, especially before the onset of rains, to avoid blockage by garbage and other human activities.

Modern-day urban flood mitigation measures include the provision of flood storage basins. Unfortunately this is impossible in Nairobi where developments are built right up to the edge of watercourses. Constrained channels thereby cause upstream flooding as there is nowhere else for the water to go.

Attempts have been made to reverse urban riparian zone encroachments , but these efforts faltered due to legal repercussions. To this day, unscrupulous developers encroach with impunity .

It’s essential that the authorities demarcate riparian boundaries and set aside buffer zones that cannot be “developed”.

• Urbanisation
• Deforestation
• Groundwater
• Water catchment
• East Africa
• Sub-Saharan Africa

Assistant Editor - 1 year cadetship

Program Development Officer - Business Processes

Executive Dean, Faculty of Health

Lecturer/Senior Lecturer, Earth System Science (School of Science)

Sydney Horizon Educators (Identified)

Texans Should Prepare For Hotter Temperatures, Greater Risk Of Fire And Flooding

Weather conditions across the Lone Star State are getting more extreme and more dangerous by the year, according to a new report from Texas A&M University professor and State Climatologist John Nielsen-Gammon.

The newly updated assessment of extreme weather in Texas draws on data from 1900-2023 to predict trends through the year 2036, and shows a significant uptick in extreme temperatures and droughts, wildfire conditions and urban flooding risks, among other changes. The report was authored by Nielsen-Gammon, a Regents Professor in the Department of Atmospheric Sciences , in collaboration with the nonprofit public policy organization Texas 2036 .

“We have national climate assessments, but they can’t do justice to Texas’ specific climate conditions,” Nielsen-Gammon said. “With this Texas-specific study, we focused on observed trends as much as possible rather than emphasizing climate model projections. The historic climate trends are part of our lived experience in Texas, and our report puts them in long-term context.”

In recent years, Texans have come face-to-face with the realities of a warming climate, sweating through record-breaking heat waves and lengthy droughts that have taken a toll on agricultural and water resources in many parts of the state.

“During the past couple of years, we’ve gone through two of the hottest summers on record,” Nielsen-Gammon said. “That has altered the trend of 100-degree days, making the increase even more dramatic than it had been. We’ve also seen new research that indicates that surface water supplies might be becoming less reliable, with increasing evaporative losses coupled with increasingly erratic rainfall.”

According to the report, those trends are expected to continue and intensify, with Texans in 2036 experiencing quadruple the number of 100-degree days compared to the 1970s and 1980s. The report also predicts a 7% increase in water lost through summertime evaporation by 2036.

All that hot, dry weather also makes the state more susceptible to wildfires, like the ones that broke out in the Texas Panhandle in February. As the report notes, the number of days with highly favorable conditions for fires to spread has already been increasing, most notably in West Texas.

Meanwhile, an uptick in statewide extreme precipitation — expected to rise 10% in intensity by 2036 compared to 2001-2020 — will likely hit the state’s eastern regions and urban centers the hardest. Parts of East Texas have already seen an increase in precipitation of 15% or more over the last century, the report notes, while a similar increase in extreme one-day precipitation has raised the risk of dangerous flooding in urban areas.

Along the Gulf Coast, rising sea levels and subsidence, or sinking of land, will increase the severity of storm surges from hurricanes, which may themselves be growing more intense due to climate change and related factors.

“Historically, the most costly natural disasters in Texas have been droughts and hurricanes,” Nielsen-Gammon said. “Hurricanes are sudden, catastrophic events, while droughts develop slowly but can affect all sectors of society. Both droughts and hurricanes are changing in complicated ways, but overall the risk from both are growing.”

Understanding these changes, as well as the distinct dangers presented by different types of natural hazards, will be crucial for communities throughout the state, Nielsen-Gammon said — from new agricultural practices in rural areas to flood control measures in cities.

“Solutions cost money,” Nielsen-Gammon said. “Drought-resilient crops tend to produce lower yields, for example. Cities can avoid new construction in flood-prone areas, but existing infrastructure is challenging to retrofit.

“A changing Texas climate presents many challenges,” he continued, “but our ability to monitor and predict those climate changes makes it easier to deal with them.”

For more information on extreme weather trends in Texas, read the full report or see the 2024 update summary published by Texas 2036.

Media contact: [email protected]

Related Stories

Sporadic Rainfall Brings Relief To Parts Of Texas

While last week's showers were a welcome sight for many, State Climatologist John Nielsen-Gammon says Texas will need “a lot more rain” to recover from its brutal summer drought.

Drought Conditions Across Texas Worsen

Texas A&M Professor and State Climatologist John Nielsen-Gammon explains what Texans can expect to see as this year’s sweltering summer draws to a close.

Summer Heat And The Texas Power Grid

The director of Texas A&M’s Smart Grid Center says he’s optimistic about the grid’s reliability as temperatures rise, but says we need to stay diligent to changing conditions.

Recent Stories

Texas A&M To Award More Than 12,000 Diplomas At Spring Commencements

Spring graduations will be celebrated May 9-11 across nine ceremonies at Reed Arena.

Helping Cats Stand Up To Pillow Foot

A Texas A&M veterinarian offers insights into the diagnosis and treatment options for a skin disease that can affect paw pads.

Shining A Light On Untapped Lunar Resources

Researchers in the Department of Aerospace Engineering have partnered with NASA Langley Research Center to design reflectors that redirect solar power to the moon’s craters.

Subscribe to the Texas A&M Today newsletter for the latest news and stories every week.

• 0 Shopping Cart

UK Floods Case Study November 2019

The UK experienced an extreme weather event in November 2019 when exceptionally heavy rainfall caused flooding in parts of the UK. Heavy downpours across large parts of northern England led to surface water and river flooding in parts of Yorkshire, Nottinghamshire, Greater Manchester, Derbyshire and Lincolnshire.

According to the Met Office, on Thursday 7th November 2019 over half of the average rainfall for the whole of November fell in parts of the Midlands and Yorkshire.

If you have images and/or videos of flooding or an eye witness account that you would be happy to share on an interactive flood impact map we are developing please send them [email protected]

What caused flooding in the UK in November 2019?

A large area of prolonged rainfall fell on parts of the UK in November 2019. Some areas experienced the whole of Novembers average rainfall over a period of 24 hours. Sheffield experienced 84mm of rainfall. The rainfall was caused by an area low pressure stalling over the UK.

Further reading/watching: 

BBC Weather Overview 

What were the effects the extreme weather in November 2019?

About 500 homes have been flooded in Doncaster with more than 1,000 properties evacuated in areas hit by the floods.

South Yorkshire Fire and Rescue said it had declared a major incident on the evening of Friday 8th November and firefighters rescued more than 40 people from the Fishlake area, near Doncaster. Residents of Fishlake said it was the first time the village had flooded in 100 years.

Empty coffins were seen floating inside the workshop of a flooded funeral parlour in the village.

Some villagers had to spend the night at a nearby pub, where staff said they had seen people crying because of the devastation.

The village church is collecting food to distribute to residents and roast dinners were delivered on Sunday to those who had remained in their homes.

Reseidents have complained that the River Don has not been dredged recently.

According to the BBC, Adrian Gill, a flood manager at the Environment Agency, said did not currently dredge the River Don “because we don’t think that’s the right thing to do” but the situation could be reviewed in the future.

Water sports enthusiast and teacher Mark Ibbotson, from Doncaster, said he, along with his 13-year-old son Logan, had rescued more than 30 people – including two babies – from a number of streets using his inflatable boat in Bentley where homes have been hit by flooding.

One of the most severely hit areas has been Bentley in Doncaster, where flooding affected many homes 12 years ago.

One resident told BBC Radio Sheffield: “The worry is our insurance policies are expensive as it is because of the 2007 floods, so now we’re all worried whether we’re going to get reinsured.”

Extensive flooding affected Rotherham , where residents were told to stay at home and not leave unless asked to do so by emergency services. Some have been taken to safety by boats.

Dozens of people were forced to spend the night in the Meadowhall shopping centre .

In Derbyshire, the River Derwent at Chatsworth reached its highest recorded level and council workers put up sandbags around Matlock and Matlock Bath, where the river was “dangerously high” .

A number of properties in Derby city centre were flooded, however, a full evacuation was not ordered as the River Derwent didn’t burst its banks to the extent emergency services believed it would.

The A52 – the main road route into Derby – was closed westbound between the city and the M1 along with a handful of smaller roads in the county.

Residents from 12 homes in Mansfield, Nottinghamshire, were unable to return home after a mudslide on Thursday led to 35 properties being evacuated .

In Nottinghamshire, residents living in mobile homes close to the River Trent in Newark were urged to move to higher ground.

On Friday, the floods claimed the life of a woman who was swept into the River Derwent at Rowsley in Derbyshire. Her body was found about two miles away in Darley Dale. She was named earlier as Derbyshire’s former High Sherriff Annie Hall .

Trains were cancelled in Yorkshire and parts of the East Midlands as rail routes were flooded.

BBC reporter Richard Cadey said some roads around Fishlake had been closed and the village was “effectively cut off because of flooding”. He said people on the ground had told him 90% of the homes there had been flooded.

The River Don, which flows through Sheffield, Rotherham and Doncaster, hit its highest recorded level at just over 6.3m (21ft), higher than it was in 2007 when it also flooded.

Sources: 

England flooding: River warnings and rail delays continue

Flooding in pictures/videos

Torrential downpours flood parts of northern England – BBC

Flooding in Yorkshire – In Pictures – The Guardian

England flooding: A tour of a flooded house in Fishlake

River Derwent Flooding – Drone Video

Helicopter captures footage of flooded South Yorkshire

What were the responses to the UK floods in November 2019?

More than 100 flood warnings were put in place across England. The Environment Agency (EA) urged people to take them seriously.

The Environment Agency took to social media to warn people about the potential impacts of flooding.

AMBER warning for flood risk today ⚠️- rain will rotate over north and north midlands bringing heavy rain on already sodden ground – take care – flood warning updates here https://t.co/K5GUW3z87V pic.twitter.com/mDDAC3uLXc — John Curtin (@johncurtinEA) November 7, 2019

The Environment Agency worked day and night to reduce the impact of flooding. The Environment Agency responded to the flood risk by working closely with police, fire and rescue, local authorities and partners to reduce the risk of flooding and keep communities safe. On the ground, Environment Agency field teams worked through the night to operate flood storage areas and pump away flood water.

A major incident was declared in South Yorkshire,

Some residents were “angry and frustrated” at Doncaster Council – claiming it had not provided sandbags early enough to prevent properties from flooding, the station reported.

Political leaders visited areas affected by floods. On the campaign trail Boris Johnson promised over £2 billion to improve flood defences.

South Yorkshire Police said it had extra officers out on patrol to “protect the evacuated areas and support those affected by the floods”.

Following a meeting of COBRA, the government’s emergency committee, Prime Minister Boris Johnson anounced the following measures :

• An extra 100 Army personnel deployed from Wednesday to support the recovery effort in South Yorkshire
• Funding for local councils where households and businesses have been affected – equivalent to £500 per eligible household
• Up to £2,500 for small and medium-sized businesses which have suffered severe impacts not covered by insurance

Six days after the heavy rain, army personnel provided support to flood-hit communities .

Environment Agency warnings

Environment Agency working day and night to reduce flood impact

How effective were the mitigation strategies introduced since the 2007 floods?

Flood defences put in place in South Yorkshire managed to significantly reduce the impact of Thursday’s floods, the Environment Agency (EA) has said.

River levels in parts of the county rose overnight to almost the same as they were in June 2007, when two people died in Millhouses and the Wicker.

Despite a major incident being declared on Thursday, the EA said the area was protected by new walls and flood gates.

The river levels around Meadowhall were high, but the EA said its defences, as well as the ones put in by Meadowhall, had lessened the damage.

Elsewhere in South Yorkshire, £3m was spent by the EA to repair and improve defences running along Ea Beck , in the villages of Toll Bar and Bentley near Doncaster.

However, people living in settlements downstream of Sheffield have complained about the impact of the recently constructed defences.  In Bentley, a low-lying neighbourhood on the north side of the River Don, forlorn terraced streets are still knee-deep in water. “You don’t have to be a hydrologist to see what’s happened,” said one man interviewed by a Guardian journalist . “Sheffield built flood defences in 2015-16. They spent about £20m protecting the lower Don. So the water has nowhere to go than the next place, Rotherham and then Doncaster.” He went on to say that residents received a “red alert” on Thursday night that there was a risk of flooding. He phoned an emergency number and requested sandbags. He was told that the council was not going to distribute them because the River Don’s banks had not been breached.

When the sandbags eventually arrived the community worked together to distribute them.

South Yorkshire flooding: Defences ‘reduce impact’

Related articles:

How do you stop flooding? 

Premium Resources

Please support internet geography.

If you've found the resources on this page useful please consider making a secure donation via PayPal to support the development of the site. The site is self-funded and your support is really appreciated.

Related Topics

Use the images below to explore related GeoTopics.

Topic Home

Next topic page, share this:.

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)
• Click to share on Pinterest (Opens in new window)
• Click to email a link to a friend (Opens in new window)
• Click to share on WhatsApp (Opens in new window)
• Click to print (Opens in new window)

If you've found the resources on this site useful please consider making a secure donation via PayPal to support the development of the site. The site is self-funded and your support is really appreciated.

Search Internet Geography

Top posts and pages.

Latest Blog Entries

Pin It on Pinterest

• Click to share
• Print Friendly

Practical Extreme Value Modelling of Hydrological Floods and Droughts: A Case Study

• Original Article
• Published: 11 January 2005
• Volume 7 , pages 5–30, ( 2004 )

Cite this article

• Kolbjørn Engeland 1 ,
• Hege Hisdal 2 &
• Arnoldo Frigessi 3  

1515 Accesses

74 Citations

7 Altmetric

Explore all metrics

Estimation of flood and drought frequencies is important for reservoir design and management, river pollution, ecology and drinking water supply. Through an example based on daily streamflow observations, we introduce a stepwise procedure for estimating quantiles of the hydrological extremes floods and droughts. We fit the generalised extreme value (GEV) distribution by the method of block maxima and the generalised Pareto (GP) distribution by applying the peak over threshold method. Maximum likelihood, penalized maximum likelihood and probability weighted moments are used for parameter estimation. We incorporate trends and seasonal variation in the models instead of splitting the data, and investigate how the observed number of extreme events, the chosen statistical model, and the parameter estimation method effect parameter estimates and quantiles. We find that a seasonal variation should be included in the GEV distribution fitting for floods using block sizes less than one year. When modelling droughts, block sizes of one year or less are not recommended as significant model bias becomes visible. We conclude that the different characteristics of floods and droughts influence the choices made in the extreme value modelling within a common inferential strategy.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Effect of climate change on the flooding of storm water networks under extreme rainfall events using SWMM simulations: a case study

An overview to flood vulnerability assessment methods.

A review on applications of urban flood models in flood mitigation strategies

Aune, B., Air temperature normals, normal period 1961–90 , Climate Report No. 02/93, the Norwegian Meteorological Institute, DNMI, Norway, 1993.

Bobée, B. and Rasmussen, P.F., “Recent advances in flood frequency analysis,” Rev. Geophys. 33, 1111–1116, (1995).

Google Scholar  

Buishand, T.A., and Demaré, G.R., “Estimation of the annual maximum distribution from samples of maxima in separate season,” Stoch. Hydrol. Hydraul. 4, 89–103, (1990).

Clausen, B. and Pearson, C.P., “Regional frequency analysis of annual maximum streamflow drought,” J. Hydrol. 173, 11–130, (1995).

Coles, S.G., An Introduction to Statistical Modelling of Extreme Values , Springer, London, 2001.

Coles, S.G. and Casson, E., “Extreme value modelling of hurricane wind speed,” Struct. Saf. 20, 283–296, (1998).

Coles, S.G. and Dixon, M.J., “Likelihood-based inference for extreme value models,” Extremes 2(1), 5–23, (1999).

MATH   Google Scholar  

Coles, S., Pericchi, L.R. and Sisson S., “A fully probabilistic approach to extreme rainfall modelling,” J. Hydrol. 273, 35–50, (2003).

Cramér, H. and Leadbetter, M.R., Stationary and Related Stochastic Processes , John Wiley and Sons, New York, 1967.

Cunnane, C., “A note on the Poisson assumption in partial duration series model,” Water Resour. Res. 15(2), 489–494, (1979).

Davison, A.C. and Hinkley, D.V., Bootstrap Methods and Their Applications , Cambridge University Press, 1997.

Davison, A.C. and Smith, R.L., “Models for exceedances over high thresholds,” J. R. Stat. Soc., Ser. B 52(3), 393–443, (1990).

MATH   MathSciNet   Google Scholar  

Dupuis, D.J., “Exceedances over high threshold: A guide to threshold selection,” Extremes 1(3), 251–261, (1998).

MathSciNet   Google Scholar  

Efron, B. and Tibshirani, R.J., An Introduction to the Bootstrap , Monographs on Statistics and Applied Probability, Chapmann & Hall, New York, 57, 1998.

Embrects, P., Klüppenberg, C. and Mikosch, T., Modelling Extremal Events for Insurance and Finance , Springer, Berlin, 1999.

FEH, Flood Estimation Handbook , Institute of Hydrology, Wallingford, UK, 1999.

Fisher, R.A. and Tippet, L.H.C., “Limiting forms of the frequency distributions of the largest or smallest member of a sample,” Proc. Camb. Phil. Soc. 24, 180–190, (1928).

Article   MATH   Google Scholar  

Frigessi, A., Haug, O. and Rue, H., “A dynamic mixture model for unsupervised tail estimation without threshold selection,” Extremes 5(3), 219ndash;235, (2002).

Førland, E., Precipitation Normals, Normal Period 1961–90 . Climate Report No. 39/93, the Norwegian Meteorological Institute, DNMI, Norway, 1993.

Førland, E., Roald, L.A., Tveito, O.E., and Hanssen-Bauer, I., Past and Future Variations in Climate and Runoff in Norway . Climate Report No. 19/00, The Norwegian Meteorological Institute, DNMI, Norway, 2000.

GREHYS, “Presentation and review of some methods for regional flood frequency analysis,” J. Hydrol. 186, 63–84, (1996).

Heo, J.-H., Salas, J.D. and Boes, C.D., “Regional flood frequency analysis based on a Weibull model: Part 2. Simulations and applications,” J. Hydrol. 242, 171–182, (2001).

Hisdal, H., Stahl, K., Tallaksen, L.M. and Demuth, S., “Have streamflow droughts in Europe become more severe or frequent?” Int. J. Climatol. 21, 317–333, (2001).

Hosking, J.R.M. and Wallis, J.R. Regional Frequency Analysis , Cambridge University Press, Cambridge, 1997.

Hosking, J.R.M., Wallis, J.R. and Wood, E.F., “Estimation of the generalized extreme- value distribution by the method of probability-weighted moments,” Technometrics 27(3), 251–261, (1985).

Kjeldsen, T.R., Lundorf, A. and Rosbjerg, D., “Use of a two-component exponential distribution in partial duration modelling of hydrological droughts in Zimbabwean rivers,” Hydrol. Sci. J. 45(2), 285–298, (2000).

Article   Google Scholar  

Klemeš, V., “Tall tales about tails of hydrological distributions. I,” J. Hydrol. Eng. 5(3), 227–231, (2000a).

Klemeš, V., “Tall tales about tails of hydrological distributions. II,” J. Hydrol. Eng. 5(3), 232–239, (2000b).

Lang, M., Ouarda, T.B.M.J. and Bobée, B., “Towards operational guidelines for over-threshold modeling,” J. Hydrol. 225(3–4), 103–117, (1999).

Leadbetter, M.R. and Rootzén, H., “Extremal theory for stochastic processes,” Ann. Probab. 16(2), 431–478, (1988).

Leadbetter, M.R., Lindgren, G. Rootzén, H., Extremes and Related Properties of Random Sequences and Processes , Springer, New York, 1983.

Madsen, H. and Rosbjerg, D., “A Regional Bayesian Method for Estimation of Extreme Streamflow Droughts,” in Statistical and Bayesian Methods in Hydrological Sciences (E. Parent, B. Bobeé, P. Hubert and J. Miquel, eds), Studies and Reports in Hydrology, UNESCO, Paris, France: 327–340, 1998.

Madsen, H., Pearson, C.P., Rasmussen, P.F. and Rosbjerg, D., “Comparison of annual maximum series and partial duration series methods for modelling extreme hydrological events 1. At-site modelling,” Water Resour. Res. 33(4), 747–758, (1997a).

Madsen, H., Pearson, C.P. and Rosbjerg, D., “Comparison of annual maximum series and partial duration series methods for modelling extreme hydrologic events 2. Regional modelling,” Water Resour. Res. 33(4), 759–769, (1997b).

Martins, E.S., and Stedinger, J.R., “Generalized maximum-likelihood generalized extreme-value quantile estimators for hydrologic data,” Water Resour. Res. 36(3), 737–744, (2000).

Martins, E.S., and Stedinger, J.R., “Generalized maximum likelihood Pareto-Poisson flood risk analysis for partial duration series,” Water Resour. Res. 37(10), 2551–2557, (2001).

Pettersson, L.-E., Fjelstad, K., Engen, T., Hydrometriske stasjoner for sanntidsdata. Vannføringsstatistikk og feltbeskrivelser. (Real time hydrometric stations. Flow statistics and catchment descriptions. In Norwegian) , Publication No. 19, NVE, Oslo, Norway, 1996.

Pickands, J., “Statistical inference using extreme order statistics,” Ann. Stat. 3(1), 119–131, (1975).

Rasmussen, P. and Rosbjerg, D., “Prediction uncertainty in seasonal partial duration series.” Water Resour. Res. 27(11), 2875–2883, (1991).

Rice, S.O., “Mathematical analysis of random noise,” Bell Syst. Tech. J. 24, 46–156, (1945).

Rosbjerg, D., Madsen, H. and Rasmussen, P.F., “Prediction in partial duration series with generalized pareto distributed exceedances,” Water Resour. Res. 28(11), 3001–3010, (1992).

Rossi, F., Fiorentino, M. and Versace, P., “Two-component extreme value distribution for flood frequency analysis,” Water Resour. Res. 20(7), 847–856, (1984).

Rootzén, H. and Tajvidi, N., “Extreme value statistics and wind storm losses: A case study,” Scand. Actuar. J. 1, 70–94, (1997).

Smith, R.L., “Maximum likelihood estimation in a class of nonregular cases,” Biometrika 72, 69–90, (1985).

Smith, R.L., “Extreme value analysis of environmental time series: An application to trend detection in ground-level ozone,” Stat. Sci. 4(4), 367–377, (1989).

Stedinger, J.R., Vogel, R.M. and Foufoula-Georgiou, E., “Frequency Analysis of Extreme Events,” in Handbook of Hydrology (D.R. Maidment, ed), McGraw-Hill, New York, 18.41, 1993.

Strupczewski, W.G., Singh, V.P. and Feluch, W., “Non-stationarity approach to at-site flood frequency modelling I. Maximum likelihood estimation,” J. Hydrol. 248, 123–142, (2001).

Tallaksen, L.M., Streamflow drought frequency analysis, Contribution to report from Exploratory Workshop on Drought and Drought Mitigation in Europe, 1–3 March 1999, JCRR-Ispra (VA), Italy (to be published in the “Natural and Technological Hazards” series of Kluwer Academic Publishers), 2000.

Tallaksen, L.M. and Hisdal, H. “Regional analysis of extreme streamflow drought duration and deficit volume. In FRIEND’97 - Regional Hydrology: Concepts and Models for Sustainable Water Management,” IAHS 246, 141–150, (1997).

Tallaksen, L. M., Madsen, H. and Clausen, B., “On the definition and modelling of streamflow drought duration and deficit volume,” Hydrol. Sci. J. 42(1), 15–33, (1997).

Tate, E.L. and Freeman, S.N., “Three modelling approaches for seasonal streamflow droughts in southern Africa: The use of Censored data,” Hydrol. Sci. J. 45(1), 27–42, (2000).

Tawn, J.A.,“Estimating probabilities of extreme sea-levels,” App. Stat. 41, 77–93, (1992).

Todorovic, P. and Rouselle, J., “Some problems of flood analysis,” Water Resour. Res. 7(5), 1144–1150, (1971).

USWRC, Guidelines for determining flood flow frequency. United States Water Resources Council, Bull. 17, Hydrl. Comm. Washington, DC, 1976.

Vogel, R.M., McHahon, T.A. and Chiew, F.H.S., “Flood-flow frequency model selection in Australia,” J. Hydrol. 146, 421–449, (1993).

Waylen, P. and Woo, M.-K., “Prediction of annual floods generated by mixed processes,” Water Resour. Res. 18(4), 1283–1286, (1982).

Woo, M.-K. and Tarhule, A., “Streamflow droughts of northern Nigerian rivers,” Hydrol. Sci. J. 39(1), 19–34, (1994).

Yevjevich, V., An Objective Approach to Definition and Investigations of Continental Hydrologic Droughts, Hydrology Papers , Colorado State University, Fort Collins, 23, 1967.

Zaidman, M.D., Keller, V. Young, A.R. Cadman, D., “Flow-duration-frequency behaviour of British rivers based on annual minimum data,” J. Hydrol. 277, 195–213, (2003).

Zelenhasic, E. and Salvai, A., “A method of streamflow drought analysis.” Water Resour. Res. 23(1), 156–168, (1987).

Download references

Author information

Authors and affiliations.

Cemagref, Lyon, France

Kolbjørn Engeland

Department of Geosciences, University of Oslo, and Norwegian Water Resources and Energy Directorate, Oslo, Norway

Hege Hisdal

Section of Medical Statistics, University of Oslo and Norwegian Computing Center, Oslo, Norway

Arnoldo Frigessi

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Kolbjørn Engeland .

About this article

Engeland, K., Hisdal, H. & Frigessi, A. Practical Extreme Value Modelling of Hydrological Floods and Droughts: A Case Study. Extremes 7 , 5–30 (2004). https://doi.org/10.1007/s10687-004-4727-5

Download citation

Received : 14 December 2000

Revised : 17 July 2004

Accepted : 23 July 2004

Published : 11 January 2005

Issue Date : March 2004

DOI : https://doi.org/10.1007/s10687-004-4727-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• generalised extreme value distribution
• generalised Pareto distribution
• maximum likelihood
• penalized maximum likelihood
• probability weighted moments
• Find a journal
• Publish with us
• Track your research

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 25 April 2024

Vegetation dieback in the Mississippi River Delta triggered by acute drought and chronic relative sea-level rise

• Tracy Elsey-Quirk   ORCID: orcid.org/0000-0002-2068-592X 1 ,
• Austin Lynn 1   nAff4 ,
• Michael Derek Jacobs 1 ,
• Rodrigo Diaz 2 ,
• James T. Cronin   ORCID: orcid.org/0000-0001-7111-6412 3 ,
• Lixia Wang 1 ,
• Haosheng Huang   ORCID: orcid.org/0000-0003-2167-3928 1 &
• Dubravko Justic 1  

Nature Communications volume  15 , Article number:  3518 ( 2024 ) Cite this article

990 Accesses

12 Altmetric

Metrics details

• Community ecology
• Natural hazards
• Wetlands ecology

Vegetation dieback and recovery may be dependent on the interplay between infrequent acute disturbances and underlying chronic stresses. Coastal wetlands are vulnerable to the chronic stress of sea-level rise, which may affect their susceptibility to acute disturbance events. Here, we show that a large-scale vegetation dieback in the Mississippi River Delta was precipitated by salt-water incursion during an extreme drought in the summer of 2012 and was most severe in areas exposed to greater flooding. Using 16 years of data (2007–2022) from a coastwide network of monitoring stations, we show that the impacts of the dieback lasted five years and that recovery was only partial in areas exposed to greater inundation. Dieback marshes experienced an increase in percent time flooded from 43% in 2007 to 75% in 2022 and a decline in vegetation cover and species richness over the same period. Thus, while drought-induced high salinities and soil saturation triggered a significant dieback event, the chronic increase in inundation is causing a longer-term decline in cover, more widespread losses, and reduced capacity to recover from acute stressors. Overall, our findings point to the importance of mitigating the underlying stresses to foster resilience to both acute and persistent causes of vegetation loss.

Similar content being viewed by others

Resilience to drought of dryland wetlands threatened by climate change

Natural recovery of a marine foundation species emerges decades after landscape-scale mortality

The impacts of tidal wetland loss and coastal development on storm surge damages to people and property: a Hurricane Ike case-study

Introduction.

Vegetation diebacks have been reported in ecosystems ranging from forests, deserts, mangroves, salt marshes, seagrass meadows, and others 1 , 2 , 3 , 4 , 5 , 6 . Dieback events have the potential to lead to longer-term population collapse, loss of communities, and ecosystem state changes 7 , although some systems can fully recover 1 . The causes of major vegetation mortality events are typically attributed to disease outbreaks, insect infestations, or acute episodic climate-mediated impacts such as severe droughts or flooding, e.g., 8 , 9 , 10 , 11 , 12 . However, the propensity for underlying chronic stresses to increase ecosystem vulnerability to acute disturbances is still poorly understood 7 , 13 , 14 . Coastal wetlands are particularly vulnerable to chronic inundation stress caused by sea-level rise. Over the last century, approximately 50% of the global coastal wetland area has been lost 15 primarily due to direct human impacts, but relative sea-level rise (RSLR) poses a compounding threat that is predicted to intensify in the future 16 , 17 , 18 . Though wetlands can be resilient to moderate rates of sea-level rise largely because of the vegetation, which traps sediment and contributes organic matter to the soil thus, allowing wetland elevation to equilibrate to increased flooding 19 , 20 , 21 , rates of RSLR that exceed thresholds of plant inundation tolerance cause wetland submergence 17 , 22 . While wetland response to RSLR is fairly well documented, the effects of acute impacts of infrequent climate-driven events such as droughts are less known, especially in areas where high rates of sea-level rise pose a chronic and escalating stress 1 .

The majority of acute wetland plant dieback events have been attributed to either excessive flooding or severe drought/low water conditions 1 , 12 , 23 , 24 , both of which can lead to the accumulation of phytotoxins in the soil 25 , 26 . Flooding has been implicated in the dieback of Phragmites australis in freshwater wetlands in Europe 12 , 27 and the U.S. Great Lakes 28 and Spartina alterniflora in a salt marsh in Texas 29 , where a lack of oxygen forces anaerobic root respiration and the build-up of reduced ions that are toxic to plants. In saline environments, or areas with sufficient sulfate, anaerobic conditions foster the accumulation of sulfide, which can lead to mortality particularly when combined with high salinity that causes osmotic and ionic salt stresses. Severe drought or low water conditions can also lead to hypersaline conditions and/or potentially soil acidification through the oxidation of reduced chemical species 23 , 26 . Nutrient-loading, herbivory, and pathogen infection have also been implicated in wetland diebacks, but almost all in combination with atypical flooding or dessication 24 , 30 , 31 . A comprehensive review of Spartina alterniflora dieback sites across the U.S. Atlantic and Gulf coasts indicates that dieback throughout the southern part of the range co-occurred with severe drought 1 . Further, dieback almost always occurred in low-elevation areas with waterlogged soils in the marsh interior 1 . The concentration of drought-related dieback incidences at low marsh elevations suggests an important role of underlying flood stress yet to be quantified and incorporated into models of wetland plant dieback.

A major constraint for identifying the causes of vegetation dieback and quantifying tolerance thresholds is the lack of field monitoring data. Here, we use a valuable 16-year dataset of field measurements (2007–2022) from a network of monitoring stations across the Louisiana coast to study a recent vegetation dieback event in the Mississippi River Delta. Vegetation dieback was first reported in the late summer of 2016 affecting Phragmites australis (Cav.) Trin. ex Steud in the modern Plaquemines-Belize “birdsfoot” delta (BFD) 32 , where the Mississippi River meets the Gulf of Mexico. However, our subsequent analysis shows that dieback began several years prior in 2012, affected multiple species, and marshes outside of the BFD. Dieback areas were initially characterized by large stands of dead stems followed by more advanced stages where only the remnant stem bases emerged from mudflats (Fig.  1 ). Initial observations of surviving Phragmites indicated an infestation of a non-native scale insect, Nipponaclerda biwakoensis (Kuwana), which specializes on Phragmites as its host plant 31 . Subsequent experimental studies show that scale infestation causes a reduction in productivity but not mortality 31 . Scale insects were first observed on Phragmites in 2017 and have since spread throughout Louisiana to neighboring states where dieback has not been reported.

a , b Aerial images in summer 2016, c oblique image of a Phragmites stand post-dieback with a healthy stand in the background, and d , e satellite imagery of Coastwide Reference Monitoring System Station 0162 lower Mississippi River Delta. Photo credits: a , b Louisiana Department of Wildlife and Fisheries, c M.D. Jacobs, d, e Google Earth.

In this study, we incorporated a rich dataset of annual vegetation cover, marsh elevation, water level, salinity, and surface elevation change time series afforded by the Louisiana Coastwide Reference Monitoring System network to (a) test empirical relationships between vegetation status and environmental conditions, principally flooding and salinity, (b) identify mechanisms likely contributing to plant mortality, and (c) quantify thresholds of tolerance and tipping points. Additionally, we tease apart the influence of short-term acute (i.e., pulse effects) and chronic (i.e., press effects) disturbances. Our analysis shows that underlying chronic flood stress predisposed low salinity marshes to an acute increase in salinity associated with low freshwater inputs and river discharge during an extreme drought. Further, by using a conceptual framework that differentiates stages and trajectories of vegetation decline and recovery, we identify thresholds for dieback, conditions for recovery, and illustrate that both periodic and long-term vegetation decline is significantly related to increases in inundation. Ultimately, these findings point to specific management recommendations to reduce chronic flood stress to improve marsh resiliency to both episodic and persistent stresses.

The Mississippi River Delta region accounts for 41% of coastal wetlands in the United States, which provides $12–47 billion in ecosystem goods and services annually 33 , 34 . However, these wetlands are extremely vulnerable to climate change and human impacts. A combination of rising sea levels, wave and storm-driven erosion, and rapid subsidence are exacerbated by human-built levees and dams causing a reduction in sediment supply, subsurface fluid extraction, and canal construction. All of these factors contribute to extremely high rates of coastal land loss up to 100 km 2 per year 35 , 36 , 37 , 38 , 39 . Accounting for land subsidence, recent rates of relative sea-level rise across the Mississippi River Delta average 1.3 ± 0.9 cm yr −1   40 , which may exceed the estimated threshold for marsh survival (0.4–0.9 cm yr −1 ) 41 , 42 . While over the last 7500 years, the Mississippi River has formed six major river deltas, and the current outlet of the main channel forms the BFD. This part of the coast is of high socio-economic importance as the Mississippi River serves as the main shipping route to the U.S. interior with over $150 billion per year in agricultural goods, chemicals, machinery, timber, coal, and steel transported through the main passes of the BFD 43 . The BFD is particularly vulnerable to RSLR because sediment input to delta marshes has been drastically reduced by upstream dams and channel protection measures 38 . Under conditions of reduced sediment supply and high rates of RSLR, vegetation loss due to a disturbance or dieback event can further hasten marsh submergence through feedbacks including a reduction in sediment trapping and surface accretion, loss of root growth and turgor, decomposition of organic matter and peat collapse 44 , 45 , 46 .

Vegetation dieback, recovery, and longer-term decline

Vegetation cover data at Coastwide Reference Monitoring System stations were examined at multiple spatial scales: (1) in the BFD ( n  = 14); (2) at all wetlands where Phragmites was present ( n  = 74); (3) at sites where Phragmites comprised >20% cover for at least one year ( n  = 33; Fig.  2a ); and (4) at a random subset of saline, brackish and freshwater marshes without Phragmites present ( n  = 30). Over the 16-year monitoring period, a significant dieback beginning in 2012, a year of significantly low Mississippi River discharge, was concentrated in the BFD and in two sites outside of the BFD where Phragmites was a dominant species, (Figs.  2 , 3 , Supplementary Fig.  1 ). Prior to the dieback, Phragmites comprised approximately 20–80% of the species coverage at dieback sites with a mix of other species making up the remaining percentage (Fig.  3 ). Coastwide Reference Monitoring System data show that the dieback began in 2012 and, in most locations, lasted five years until 2017. Dieback was characterized by an almost complete loss of Phragmites and a significant decline in other species (Fig.  3 ). Dieback sites in the BFD experienced only a partial recovery in 2017 while the two sites outside of the BFD had a full recovery to pre-dieback vegetation cover by 2017 (Fig.  3 ; Supplementary Fig.  1 ). Over the 16 years, marshes in the BFD also experienced a longer-term decline in species richness, total vegetation cover, and Phragmites cover concurrent with a shift to more flood-tolerant species, specifically Colocasia esculenta (Fig.  3 ). Stands with the highest Phragmites cover before the dieback in the BFD experienced a progressive decline with 79 ± 5% Phragmites coverage in 2011, 47 ± 9% in 2017 and 25 ± 3% in 2022, representing an average decline of 68% over the 16-year monitoring period (Figs.  1d, e and  3 ; Supplementary Fig.  1 ).

a Location of marshes that experienced dieback, no dieback, or lacked data over the dieback period, 2012–2017 based on Coastwide Reference Monitoring System data where Phragmites australis was a community dominant species (>20% cover), Google Earth, 2023. b seasonal Mississippi River discharge, Baton Rouge, LA, 2007–2022.

The black scatter line represents the total annual vegetation cover (mean ± standard error) across sites ( n  = 16). The stacked bars are mean species cover data for all dominant and frequent species. Source data are provided as a Source Data file.

Dieback sites experience greater flooding and lower salinities than non-dieback sites

All dieback stations were dominated by Phragmites yet twenty stations dominated by Phragmites located farther west along the Louisiana coast did not experience dieback (Fig.  2a ; Supplementary Fig.  1 ). To investigate whether flooding differed among dieback and non-dieback sites, we used daily marsh inundation, a measure of water level relative to the marsh surface, using a combination of water level (referenced to NAVD88) and marsh elevation (referenced to NAVD88) at each Coastwide Reference Monitoring System station (see “Methods” section for details). We tested the strength of relationships between inundation, salinity, and vegetation during different time periods over the growing season and found that flooding and salinity dynamics from May through July had the most profound influence on summer vegetation cover, as compared to later in the growing season (August through October) or averaged over the entire growing season (March through October), likely due to this being a critical period for plant growth.

A comparison of marsh inundation and salinity May through July from 2008 through 2021 between dieback and non-dieback sites revealed that (1) dieback sites are persistently subject to greater flood depths and lower salinities than non-dieback areas; (2) dieback areas had exceptionally high mean daily and maximum salinity in 2012; (3) marsh inundation is increasing over time at all locations; and (4) tolerance to high salinity is greater under less flooded conditions as reflected by the high vegetation coverage and low dieback in saltier but less flooded sites (i.e., non-dieback sites; Fig.  4 ). Across the time series, marsh inundation depth averaged 15 cm higher in the dieback areas than non-dieback areas (Fig.  4a ). Maximum flood depths May through July, which represent occurrences of storm flooding and, in the BFD, river discharge, were similar between dieback and non-dieback areas indicating that differences are due to average conditions, not extreme events (Fig.  4c ). Salinities are typically very low, less than 1 psu in dieback marshes because of the proximity of most sites to the Mississippi River, but average between 3 and 5 psu and up to 10 psu in the non-dieback areas (Fig.  4b, d ). In 2012, mean and maximum salinity in the dieback areas were exceptionally high, averaging 4.5 and 13 psu, respectively May–July.

a Mean marsh inundation, b mean daily maximum salinity, c mean maximum marsh inundation, and d mean maximum salinity May through July 2008–2021 across dieback ( n  = 16) and non-dieback ( n  = 20) sites. Dieback occurred in marshes where Phragmites comprised >20% of the vegetation community for at least one year of monitoring therefore this comparison was focused on marshes composed of >20% Phragmites for at least one year. Values are means ± SE.

Though dieback marshes have experienced greater flood depths over the last 16 years, marsh inundation has increased at all sites (Fig.  4a ). As marsh flooding increased in non-dieback areas, salinities have declined suggesting greater freshwater inputs (Fig.  4a, b ). However, with an increase in inundation across the region, the vulnerability to acute disturbances such as salinity pulses from droughts or storms is amplified both spatially and temporally.

Acute dieback caused by drought-induced salinity incursion and exacerbated by tropical cyclones

Vegetation dieback coincided with the second most severe U.S. drought on record 47 , 48 and Tropical Storm Debby in late June 2012 49 causing elevated salinities throughout the early part of the growing season, May–July (Figs.  3 , 4 , and 5a ). The drought resulted in low Mississippi River discharge and prolonged high salinities during the spring and summer. Tropical Storm Debby affected the dieback region for several days beginning approximately June 20, 2012. Our analysis shows that high background salinities associated with the drought likely facilitated the unusually high salinities during this early-season tropical storm. Except for Tropical Storm Debby, all early tropical cyclones in June and July across the 16-year time series were freshwater flooding events (Supplementary Fig.  4 ). By parsing out the drought- and storm-related flooding and salinities (see “Methods” section), we find that drought-related salinities persisted for an extended period May through July with over 20 days of >5 psu and over 10 days with salinity exceeding 10 psu (Supplementary Fig.  5 ). The close proximity of the Loop Current to the mouth of the Mississippi River in July also brought high salinity water near the coast (Supplementary Fig.  6 ). Marsh inundation in May 2012 was lower than any other year from 2008 to 2022 ( p  < 0.01), when the Mississippi River discharge was very low (Fig.  2b ) and 24% of the contiguous U.S. was experiencing severe to extreme drought 47 . However, despite the relatively low water, dieback marshes averaged 13 cm greater water depths than non-dieback marshes and remained saturated with high salinity water (Fig.  4a, b ). Tropical Storm Debby exacerbated these conditions by causing 45 cm of marsh inundation and 17 psu salinity across monitoring locations for several days (Fig.  5b , Supplementary Figs.  3 and 4 ). Hurricane Isaac followed the initial detection of the dieback in late August and caused saltwater flooding of almost 1 m across the dieback marshes with maximum salinities averaging 25 psu, greater than any other tropical cyclone in the 16-year time series (Fig.  5b , Supplementary Fig.  4 ). The high flooding and unprecedented salinities during Hurricane Isaac likely contributed to the protracted 5-year impact on the vegetation.

a Mean daily salinity during non-storm periods, May–July, b mean maximum salinity during Jun–July and August–October periods, c generalized linear model of mean percentage of time the marsh is flooded over the growing season and d generalized linear models of mean daily and monthly maximum marsh inundation depth over the growing season across 14 Coastwide Reference Monitoring System stations in the BFD ( n  = 14, ±standard errors). Source data are provided as a Source Data file.

Thresholds for dieback and conditions related to decline and recovery

Determining empirical relationships between vegetation status and environmental conditions can be complicated by hysteresis in vegetation responses to ephemeral disturbances. For example, vegetation loss can be prolonged even after the stressor is removed and conditions have been restored due to biological lags in regeneration capacity, recolonization, and growth 50 . Because vegetation response may be delayed or on a trajectory of decline or recovery trailing or disparate from the immediate environmental conditions, plant status may not directly reflect abiotic conditions at a particular time. Here, we used a conceptual framework that classifies distinct vegetation states based on relative percent cover over time and trajectories from previous years modified from a model describing hurricane impacts on coral reef declines 51 . This framework was used to compare environmental conditions among states (Peak, Decline, Dieback, Impact, Recovery, and Resumption) and test relationships between vegetation cover and environmental variables within states (Fig.  6a, b ). For each Coastwide Reference Monitoring System station where dieback occurred, total vegetation cover each year was classified as one of six vegetation states. The Peak state occurred during years with the highest (90th percentile) vegetation coverage for each site. The Decline state was characterized by years with a reduction in cover which was lower in magnitude than the Dieback state. Vegetation cover during Declines averaged 78 ± 8%, (a 28% decline from Peak coverage; p  < 0.01), and 51 ± 13% during the Dieback. Phragmites cover ranged from 4–60% during Declines and 0–3% during the Dieback. Distinguishing lower magnitude vegetation declines from the dieback allows us to examine the potential for declines and dieback to be affected by different drivers. An Impact state was characterized by years with low vegetation cover following a Decline or Dieback when abiotic conditions (i.e., inundation and salinity) were restored. The Recovery state occurred in years with an increase in cover following Declines, Diebacks, or Impact periods, and Resumption represented a stasis following a Recovery that did not reach Peak cover (Fig.  6a ). It is important to acknowledge that our system is also experiencing a declining baseline (Figs.  3 , 6b ) but for purposes of this analyses, we included only conditions during the highest peak coverages using the stable population framework (Fig.  6a ).

a Conceptual framework of vegetation states based on trajectories of vegetation cover with a stable steady-state baseline modified from ref. 51 , b conceptual framework of vegetation states based on trajectories of vegetation cover with a declining baseline modified from ref. 51 , c relationship between vegetation cover and hydroperiod during the growing season for different vegetation states, d relationship between vegetation cover and duration of salinity >10 psu during the growing season for different vegetation states. Data are means ± standard error.

Peak vegetation cover averaged 109 ± 4% (Fig.  3 ; Fig.  7 ; p  < 0.01). Vegetation cover measured at Coastwide Reference Monitoring System stations are the sum of individual species covers that include under-, mid-, and overstory canopy and vine species and therefore can exceed 100%. Peak cover occurred when the marsh surface was flooded approximately 50% of the time over the growing season and salinity averaged less than 2 psu and was above 10 psu for less than 10 days (Fig.  6c, d ; Supplementary Fig.  7 ). The Dieback was characterized by a high magnitude of vegetation loss and, for the dominant species, Phragmites , 80 to 100% lower coverage (Figs.  3 , 6 ). Marsh inundation depth and percent time flooded was similar during Peak and Dieback states and higher during the Impact, Decline, Recovery and Resumption periods (Fig.  6c, d ; Supplementary Fig.  7 ). During the dieback, average salinity over the growing season was over four times greater than during other states (F 5, 173  = 120.5, p  < 0.0001) and salinities above 10 psu lasted for an average of 27 days as compared to 5 days during Peak coverage (Fig.  6d ; Supplementary Fig.  7b ). Lesser magnitude Declines occurred both before and after the high magnitude Dieback primarily in 2008, 2010, and 2020 and over the long-term (Fig.  3 ). Impacts rarely occurred following a decline but were consistent across sites following the Dieback. Vegetation cover during the Impact state was statistically similar to the Dieback and for Phragmites averaged 1.6 ± 0.6%. Recovery states occurred following Decline, Dieback, and Impact states notably in 2009, 2011, 2017, and 2021. Resumptions occurred in 2018, 2019, and 2022. Using these population states, we examine relationships and threshold dynamics across and within states (Fig.  7 ).

a 3-D mesh interpolation of the combined effects of marsh inundation and maximum daily salinity over the growing season on Phragmites dieback (includes sites with initial cover above 50%), b relationship between maximum marsh inundation (May–July) and change in percent cover during declines based on simple linear regression c 3-D mesh interpolation of the combined effects of marsh inundation (May–July) and daily maximum salinity (August–October) on Phragmites recovery. Note y-axis is reversed to show the low cover at high salinity and low inundation.

Thresholds and relationships

Focusing on Phragmites cover dynamics and tolerances, we found no threshold marsh inundation level that characterized the transition between peak cover and dieback, which occurred during a period of relatively low water ( p  > 0.05). The transition from peak to dieback cover was described by a significant sigmoidal relationship with mean daily maximum salinities over the growing season (sigmoidal model: y = a/(1 + e^(−(x − c)/b); where a  = 147, b = −1.78, and c  = 4.91; Adj R 2  = 0.54, p  = 0.0003)) indicating a threshold value of 7.2 psu (Fig.  7a ). This salinity threshold represents an average condition over the growing season when dieback was initiated but not an absolute threshold for mortality.

The magnitude of Phragmites cover loss during interannual Declines (primarily in 2008, 2010 and 2020), which was less than that during the dieback, was best explained by maximum marsh inundation May through July indicating that early-season extreme flooding from high river discharge and/or tropical cyclones reduces plant density during that summer (Fig.  7b ). The most severe and spatially consistent decline in cover occurred between summer 2019 and 2020 following excessive freshwater flooding from Tropical Storm Cristobal in June 2020 and five succeeding hurricanes, which caused elevated water levels and salinities July through October, and likely contributed to a limited recovery of Phragmites (<10%) by summer of 2021 (Supplementary Fig.  1 ). Overall, maximum marsh inundation May–July explained approximately 60% of the variation in vegetation declines (Fig.  7 b ). And while declines in cover were also statistically related to mean inundation averaged across the growing season, the relationship was strongest with maximum inundation May through July. Phragmites recovery from dieback and declines occurred when salinities were low with an optimal flood depth between 20 and 30 cm from May through July based on a parabolic relationship ( R 2  = 0.19, p  < 0.05) indicating that relatively moderate early-season flooding increased the percent of recovery (Fig.  7c ).

Longer-term vegetation loss related to an increase in inundation

Local RSLR across the BFD, calculated using water level time series data, has been increasing at a rate of 2.7 ± 0.1 cm yr −1 over the last 16 years ( R 2  = 0.66, P  < 0.01, y  = 2.66x + 0.04;) due to a combination of subsidence and eustatic sea-level rise. However, marsh inundation depends on marsh surface elevation and the rate of marsh elevation change. Marsh elevation change rates in the BFD, measured biannually using surface elevation tables, range from 0.3 to 5.8 cm yr −1 and average 2.1 ± 0.4 cm yr −1 . Marsh inundation, which incorporates water level, surface elevation change, and surveyed marsh elevations, across the delta during the growing season has been increasing at a rate of 0.75 cm yr −1 since 2008 (Fig.  5d ). Monthly maximum flood depth has also increased but at a much higher rate of 1.2 cm yr −1 . Further, the percentage of time that the marshes are flooded (i.e., the percentage of days that the mean water level was above the marsh surface), has increased from an average of 43% to 75% (Fig.  5c ).

The temporal trends in vegetation decline over the 16-year time period correlate to increases in marsh inundation and hydroperiod (ANCOVA: R 2  = 0.34, p  < 0.0001; Figs.  3 , and 5c, d ) and further, total vegetation cover declines significantly with an increase in marsh inundation depth over the growing season (Fig.  8 ). Excluding the Dieback state when water levels were markedly low, vegetation cover declined with an increase in inundation across all population states indicating that the degree of Decline, Recovery, and Resumption and absolute extent of vegetation cover are governed largely by flooding. The long-term trends of increasing marsh inundation in the BFD coincide with shifts in the vegetation community to fewer, more flood-tolerant species (Fig.  3 ). Species richness averaged 11–14 species before 2012 and has ≤7 species over from 2019–2022. The decline in species richness is accompanied by a reduction in robust emergent species such as Phragmites , Sagittaria platyphylla , Spartina alterniflora , and Spartina patens and an increase in the flood-tolerant Colocasia esculenta (Fig.  3 ).

Generalized linear model based on linear regression of mean marsh inundation and vegetation cover. The shaded area represents 95% confidence interval from the mean. Source data are provided as a Source Data file.

The Mississippi River Delta region is one of many coastal areas around the world vulnerable to wetland loss due a combination of human management and climate-driven changes 34 , 35 , 36 , 37 . Vegetation dieback in this system was associated with an acute increase in salinity particularly during the early to peak growing season. Here we show that recovery from acute causes of vegetation loss can occur but is tempered by underlying chronic stresses to the system. Both short- and long-term declines in vegetation cover and richness were associated with high levels of marsh inundation because of marsh subsidence and sea-level rise. Therefore, managing the source(s) of chronic stress, in this case, RSLR, is imperative to facilitate resilience to both short-term disturbances as well as the underlying causes of longer-term decline.

Though the focus of this study is on the impacts of salinity and inundation on vegetation loss, there are certainly other anthropogenic changes occurring in the Mississippi River Delta that are likely important stressors and threats to plants. Nutrient inputs from agricultural, sewage and urban-development runoff leading to eutrophication 52 , 53 , 54 , toxic heavy metals such as copper, iron, and zinc 55 , frequent oil spills 56 , invasive plant species (e.g., C. esculenta 57 , 58 ) and rising temperatures 59 are all potential stressors to plant communities in coastal marshes. In 2010, two years prior to the documented dieback, the BFD and other areas along the northern Gulf coast were impacted by the BP-Deepwater Horizon oil spill. In marshes exposed to oiling, vegetation mortality was concentrated at the edge where oil coverage was greatest and, in some locations, recovery of marsh grasses in previously oiled areas occurred within 18 months 60 . An analysis of the BFD marshes showed little impact of oil on the vegetation attributed to relatively low oil residence time due to river discharge and limited oil deposition on the soil 61 . Though there is the potential for the oil spill to have impacted dieback areas two years before the dieback, the strong temporal correlations shown here suggest that salt intrusion and underlying flood stress were important. Nutrient concentrations, particularly nitrate concentrations, have increased in the Mississippi River over time 62 and high nutrient availability has been hypothesized to contribute to Phragmites dieback in other systems 12 , 27 . Nutrients cause an increase in organic matter production which, under anaerobic conditions, can intensify the production of organic acids and sulfides 27 and while not directly tested here may also contribute to vegetation stress in the BFD. While we show that recovery of vegetation is influenced by RSLR, it is not the only chronic stressor likely to slow recovery. Introduction of the Roseau cane scale ( N. biwakoensis ) into the Mississippi River Delta, a specialist of Phragmites and first reported in 2017 32 , is clearly slowing recovery of dieback sites. In a common-garden experiment, low to moderate scale densities (≈20 per m of the stem) reduced end-of-season biomass by 28% relative to plants with few scales (4 per m of stem). Peak densities of scales, occurring in late summer, have consistently exceeded 150 per m of stem since 2017 32 . Scale densities among Phragmites stands in the Mississippi River Delta are negatively correlated with the normalized difference vegetation index (NDVI); a measure of standing plant biomass (I. A. Knight, G. Suir and R. Diaz, unpublished data). All of these stressors have the potential to interact, elevating their chronic effects to an acute level.

Greater marsh inundation affected the susceptibility to vegetation dieback and while not the cause of acute dieback, caused interannual and longer-term population and diversity declines. Vegetation dieback beginning in 2012 was concentrated in the BFD where both rates of RSLR and marsh elevation change are relatively high 40 . But, as compared to non-dieback sites farther west, the BFD marshes are lower in elevation and affected by both river and Gulf of Mexico hydrodynamics (Supplementary Fig.  8 ). High flood depths averaged across the growing season were strongly related to low vegetation cover, while the magnitude of annual declines in cover was influenced by early-season high maximum marsh inundation. Elevated water levels were attributed to dieback of Phragmites in Europe 12 , 63 , several native marsh plant species in the southeastern United States 56 , 64 , and mangroves on the Indo-Malayan coast 65 . Our data show that high river discharge and marsh inundation in the BFD induces an immediate short-term loss of vegetation, yet the chronic increase in average marsh flooding is contributing to a slower longer-term decline. Recovery of Phragmites following declines and dieback was dependent on an optimal moderate level of spring flooding and in restoration trials, Phragmites plantings in dieback areas had low survivorship in areas of greater water depth 66 . This relationship with flooding is similar to responses of other wetland plants to experimental sea-level rise 19 and supports findings in salt marshes where vegetation recovery following a disturbance slows with an increase in seawater inundation 67 .

Acute dieback occurred during an extreme drought with the proximate cause of saltwater intrusion. Most of the dieback sites were located adjacent to the Mississippi River where low discharge caused upriver transport of the salt wedge. While this dieback was associated with an extreme drought, other river management activities that lower freshwater discharge (e.g., upstream dams and reservoirs, diversion of river water) can also cause low water conditions in the river. To accommodate larger ships, the U.S. Army Corps of Engineers has deepened the Mississippi River Ship Channel to a minimum of 53 feet. Deep navigation channels can facilitate the upstream movement of saline water particularly during low river discharge.

The dieback targeted Phragmites -dominated marshes. Marshes near dieback sites dominated by other species including more salt-tolerant species, were relatively stable during the dieback period. Coastal plant diebacks tend to impact single species. Examples include diebacks of Spartina alterniflora , Phragmites australis, and the seagrass Thalassia testudinum that all typically grow in large monocultures 1 , 6 , 12 . Seagrass dieback in Florida Bay, for example, targeted dense stands of Thalasia , while sympatric species remained unaffected. Turtlegrass was susceptible to dieback because of its growth habit of forming dense meadows where high biomass and respiratory demand for oxygen led to phytotoxic sulfide incursion 68 . Although it is not entirely clear why Phragmites -dominated marshes were exposed to dieback in the Mississippi River Delta, a similar dynamic to Thalassia may be occurring and previous studies have implicated the accumulation of organic acids and sulfides as the products of anaerobic decomposition in the rhizosphere 12 , 27 , 69 . This combined with high salinities, which can further facilitate the generation of sulfides, can be lethal. Salinities that exceed 15 psu in the root zone have been associated with Phragmites dieback prior to the onset of flowering across 27 habitats along the eastern and western coasts of Denmark 70 . We found a dieback threshold of 7 psu averaged across the growing season that represents significant time periods of relatively high salinity (e.g., 20 days ≥ 5 psu, 10 days ≥ 10 psu) during the drought and extreme salinity and flooding pulses during subsequent tropical cyclones. However, relatively high salinities and low water levels in non-dieback marshes in the present study provide further support that Phragmites can tolerate high salinities as long as the sediments are relatively oxidized 71 .

Implications

As we write this paper, drought conditions have caused two consecutive years of historically low water levels in the Mississippi River. Based on historic temperature and precipitation trends in the Mississippi River watershed, extreme summer droughts are predicted to become more frequent in the future in response to climate warming 72 . The loss and potentially limited recovery of marsh vegetation from a prolonged salinity intrusion event and chronic flood stress, as we have documented, can have a number of ecosystem-level consequences including the loss of breeding and foraging habitat for birds 73 , 74 , muskrats, and the American alligator 75 , changes in soil biogeochemistry 27 , 76 , 77 , lower accretion rates and soil strength 78 and thus a reduced ability to keep pace with sea-level rise 21 . Vegetation plays a pivotal role in wetlands by stabilizing channel banks, facilitating sediment deposition, contributing organic matter to soil and oxygenating the rhizosphere. The consequences of mass mortality may lead to peat collapse and conversion to mudflat through the collapse of air-filled rhizomes and decomposition of organic matter. Lower elevation, more flooding, and anoxic soil conditions may limit the recovery of marsh vegetation resulting in a state change from emergent marsh to mudflat/submerged habitat without a large input of allochthonous sediment 45 (  Supplementary Figs.  9 – 13 ) . The Mississippi River Delta is suffering from catastrophic rates of wetland loss and marshes of the BFD are disappearing rapidly. Our findings suggest that a reduction in chronic inundation through the delivery of sediment (e.g., enhanced flow through channels and connectivity to marshes, thin-layer placement) to increase marsh elevations will reduce the vulnerability of marsh vegetation to episodic disturbances and enhance recovery following disturbances.

The dieback was initially reported as affecting Phragmites australis in the birdsfoot delta (BFD). To examine the scale and magnitude of the vegetation dieback, we examined the time series of species percent cover data collected annually during peak growing season since 2006 or later at Louisiana Coastwide Reference Monitoring System Stations. The Coastwide Reference Monitoring System network is comprised of 391 monitoring stations across swamp, freshwater, intermediate, brackish, and saline wetlands with data collection beginning in 2006 and continuing to the present day. To test whether the dieback is reflected in the annual vegetation cover data at Coastwide Reference Monitoring System stations, we examined the 16-year time series of species cover data at Coastwide Reference Monitoring System stations in the BFD ( n  = 14), in all marshes where Phragmites was present ( n  = 74), and at a random subset of saline, brackish and freshwater marshes without Phragmites present ( n  = 30). Dieback was detected at sites with a relatively high coverage of Phragmites (>20% for at least one year since 2007) and was concentrated in the BFD. Two sites outside of the BFD where Phragmites was dominant also experienced dieback (Fig.  2b ). Summary annual data used for this study are provided as Supplementary Data  1 .

Subsequent analyses were conducted to examine (1) differences between dieback and non-dieback marshes in hydrology (i.e., marsh inundation and percent time flooded – see below for detail) and salinity over time; (2) annual differences in hydrology and salinity in dieback marshes; and (3) relationships between vegetation cover and hydrology and salinity during peak, dieback, declines, recoveries, and other population states and across the time series to determine potential thresholds of tolerance. For these analyses, we used Coastwide Reference Monitoring System annual vegetation percent cover, marsh elevation (NAVD88), marsh surface elevation change, hourly water level, and hourly salinity data. All raw data used for this study are publicly available 79 .

Each Coastwide Reference Monitoring System station is comprised of a 200-m × 200-m area where marsh elevation, marsh surface elevation change are measured, vegetation plots are located and soil cores are collected and other station measures (e.g., periodic plant biomass) are taken. At each station, continuous water level, salinity, and temperature data are recorded hourly from a nearby channel or bay. Detailed information on the methodology of Coastwide Reference Monitoring System station establishment and data collection can be found in ref. 80 . Table  1 below shows a brief outline of methods for vegetation and hydrology used in this study.

Vegetation cover

Vegetation measurements at Coastwide Reference Monitoring System stations are collected during the summer, approximately June through August, each year. At each station, percent cover data is collected within ten permanent 2 m × 2 m plots along a 283 m transect using the Braun-Blanket cover class method: 0%, <1% 1–5%, 5–25%, 26–50%, 51–75%, 76–100% 81 . The site mean of the mid-point of each cover class for each species is reported for each year. Total vegetation cover of each station and cover of each layer (e.g., canopy and understory) is estimated between 0 and 100 percent. Therefore, the sum of the species cover (i.e., total cover) may exceed 100 percent because of overlapping canopies.

Assessment of percent cover data by basin illustrated that acute dieback began in 2012 and lasted until 2017 and was concentrated in the BFD ( n  = 14). Unfortunately, percent cover data was not recorded for some of the BFD Coastwide Reference Monitoring System stations during the dieback period (2012–2017). Field notes indicate dense Phragmites stands at the marsh edge and that the coverage in the interior plots was assumed to be 100%. However, vegetation condition at the marsh edge does not necessarily reflect the condition of vegetation in the marsh interior. All other stations in the BFD showed a significant decline in cover that lasted approximately 5 years. All but one site in the lower delta has experienced a significant longer-term decline in coverage over the 16 years.

Phragmites haplotypes

The vegetation on the outer fringe of the BFD consists of virtually monospecific stands of 4–5-m-tall Phragmites 82 with a diversity of brackish to freshwater marsh plants also present 83 . In 1968, Phragmites comprised 11, 57, and 39% of the salt, brackish, and fresh marsh vegetation respectively 84 , and since then have been considered the most common dominant marsh species in the BFD 33 . Four distinct haplotypes of Phragmites australis occur along the Gulf coast from Florida to Texas 85 . Land-type (I2), occurring inland primarily along roadside ditches; Delta-type (M1), the most widespread haplotype in Louisiana marshes, EU-type (M), the notorious invasive haplotype in many areas around the world, but only found in localized patches in marshes of Louisiana, and the Greeny-type (AI) – uncommon and in localized patches. Coastwide Reference Monitoring System vegetation data does not distinguish Phragmites by haplotype, and while both EU- and Delta-types are present in marshes of the BFD, the Delta-type is dominant and more widespread than the EU-type.

Hydrology and salinity

Water level and salinity datasets for this study cover a time period from approximately July 2007 through December 2022. Because only the latter part of 2007 data were available, 2008 through 2022 were used for analyses of relationships between hydrology, salinity, and Phragmites cover. Water level are referenced to the NAVD88 datum. Water level and salinity data are collected hourly, and for this analysis mean daily water level (i.e., average of the 24 hourly water level values) and maximum daily salinity (i.e., maximum of the 24 hourly salinity observations) were used.

Marsh elevations

Station elevations were surveyed in 2014 using an RTK-GPS at the RSET, the sonde and staff gauge, and the marsh surface. The elevations were relative to NAVD88 using Geoid 12A.

Marsh elevation change

Rod Surface Elevation Tables (RSET) were established at each Coastwide Reference Monitoring System station for measuring marsh elevation change over time. The RSET methodology 86 consists of stainless steel rods driven vertically into wetland sediments to the point of refusal and secured in concrete bound with a PVC-pipe collar. During each site visit (1–2 times yr −1 ), an RSET arm is attached to the rod, and in each of four directions, nine pins are lowered to the wetland surface, for a total of 36 pin height measurements per site visit. Marsh elevation change is calculated by taking the mean difference between subsequent pin heights and initial pin heights across the 36 measurements. An annual rate of change is calculated for each station using a linear trend in elevation change.

Marsh inundation

For each Coastwide Reference Monitoring System station, we calculated “daily marsh inundation” to represent water level relative to the marsh surface using the daily mean water level (NAVD88), initial marsh elevation in 2014 (NAVD88), and marsh surface elevation change over the 16-year period. Surface elevation change is measured biannually using RSETs 80 . Incorporating marsh elevation change is essential for calculating marsh inundation in areas experiencing shallow subsidence and accretion that result in marsh elevation change over the period of study. Linear interpolation was used to estimate daily marsh elevations between elevation change measurements. Daily marsh inundation was calculated over the 15-year time series by subtracting daily marsh elevation from the mean daily water level for each station. Positive values indicate water level is higher than the marsh elevation and, thus, the marsh surface is submerged.

Surface water salinity

Daily maximum salinity is defined as the maximum value of the 24 hourly observations in a day, which was typically within 2 psu of daily mean salinity, was used for all analyses as maximum salinities were considered most relevant for assessing contributions to plant stress and loss of cover.

Daily mean inundation and daily maximum salinity values were averaged across Coastwide Reference Monitoring System stations over specific time periods (i.e., growing season: March–October, early growing season: May–July, and late growing season: August–October) for each year relevant to seasonal hydrology and plant phenology. May through July represents mid-peak growing season and a time period of high to low river discharge, generally freshwater conditions, and early tropical cyclone activity. August through October represented late growing season, low river discharge, and higher salinities, and late tropical cyclone activity. Including data from colder months (January–April and November–December) in the analyses generally lowered the strength of significant relationships between abiotic conditions and vegetation response.

Marsh inundation and salinity metrics were also calculated with and without tropical cyclone impacts. Cyclone impacts were removed by first compiling information on all Gulf of Mexico cyclones over the 2007 through 2021 time period using archived tropical cyclone reports (National Hurricane Center, NOAA) then identifying corresponding anomalous patterns of water level and salinity at each Coastwide Reference Monitoring System station. Datasets were partitioned into those with and without storm-related water levels and salinity.

To test differences in hydrology and salinity between areas that experienced dieback and those where dieback did not occur, we focused our analysis on dieback and non-dieback Coastwide Reference Monitoring System stations where Phragmites cover was greater than 20% for ≥1 years of the 16-year time series ( n  = 36). Stations were classified as freshwater and intermediate salinity marshes. We tested differences between dieback and non-dieback marshes in mean daily marsh inundation, mean maximum marsh inundation, mean daily maximum salinity, and mean maximum salinity during the early growing season May through July, a time period where abiotic variables were most strongly related to vegetation response.

Vegetation state conceptual framework for testing empirical relationships

To test relationships between vegetation cover and marsh inundation and salinity, vegetation cover was classified into one of six vegetation states: Peak, Decline, Dieback, Impact, Recovery, and Resumption that reflect the relative cover and the trajectory from previous year’s (Fig.  6 ). Except for Peak coverage, the classification of the other states was dependent on the prior year’s state. Peak cover was classified as the highest 90th percentile of vegetation coverage at each site across the 16-year time series. Declines in cover were considered lower magnitude reduction in cover from the previous year in years other than the dieback years (2012 and 2013). Dieback in contrast was a high magnitude loss of cover in 2012 and 2013. Impact periods were a following a decline or dieback. Recovery was an increase in cover following a decline, dieback, or impact state. Resumption was a stasis following a recovery but lower cover than during Peak states. This framework was used to test differences in marsh inundation and salinity (daily mean and maximum) over the growing season, March through October among states using analysis of variance (ANOVA) and test relationships between vegetation cover and environmental variables within states using regression analysis. For these analyses, marsh flooding and salinity dynamics over the entire growing season were included to capture the strong seasonal variation and relative influence of freshwater river discharge in the spring and higher salinities associated with lower discharge, coastal set-up, and tropical storms in the late summer, early fall (Supplementary Fig.  8 ).

Relationship between marsh inundation and salinity and Phragmites cover

Multiple linear regression was used to examine the influence of marsh inundation and salinity on the percent cover of Phragmites during specified time periods (i.e., May–October, May–July, and August–October) and the population states of Phragmites . Additionally, we used a Welch’s ANOVA (unequal variances) to test differences among states in marsh inundation and salinity metrics 87 . The Games-Howell test was used for all multiple pairwise comparisons.

To test the potential for thresholds, we conducted sigmoidal regression analyses 88 , 89 that included relative percent cover as the dependent variable and salinity and marsh inundation during important seasonal time periods as the independent variables. Relative percent cover was used instead of absolute percent cover to account for differences in initial coverages of Phragmites across sites. For the sigmoidal analyses, we used the inflection points of the sigmoidal equations to identify thresholds in salinity and inundation for Phragmites cover. Sigmoidal regression, exponential limited growth regression, and Spearman rank correlation analyses were conducted in Sigma Plot Version 13.0 (Systat Software, San Jose, California, USA). Linear regression was used to examine the relationship between vegetative cover, marsh inundation, and salinity across sites and time. An Analysis of Covariance was used to test the decline in vegetation cover over time and the influence of marsh inundation. Data associated with the dieback were removed for this analysis. All statistical analyses were conducted using JMP SAS Pro V.16 (SAS Institute, Cary, North Carolina).

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

Source data are provided as a Source Data file. Annual summary data are available as a Supplementary Data  1 file and all raw data are publicly available from the Coastal Protection and Restoration Authority (CPRA) of Louisiana, via the Coastwide Reference Monitoring System program and can be retrieved from the Coastal Information Management System (CIMS) database ( http://cims.coastal.louisiana.gov ).  Source data are provided with this paper.

Alber, M., Swenson, E. M., Adamowicz, S. C. & Mendelssohn, I. A. Salt marsh dieback: an overview of recent events in the US. Estuar. Coast Shelf Sci. 80 , 1–11 (2008).

Article   ADS   Google Scholar  

Allen, C. D. & Breshears, D. D. Climate-induced forest dieback as an emergent global phenomenon. Eos Trans. AGU 88 , 504–504 (2007).

Hamerlynck, E. P. & McAuliffe, J. R. Soil-dependent canopy die-back and plant mortality in two Mojave Desert shrubs. J. Arid Environ. 72 , 1793–1802 (2008).

Rizzo, D. M. & Garbelotto, M. Sudden oak death: endangering California and Oregon forest ecosystems. Front. Ecol. Environ. 1 , 197–204 (2003).

Article   Google Scholar  

Duke, N. C. et al. Large-scale dieback of mangroves in Australia’s Gulf of Carpentaria: a severe ecosystem response, coincidental with an unusually extreme weather event. Mar. Freshw. Res. 68 , 1816–1829 (2017).

Zieman, J. C., Fourqurean, J. W. & Frankovich, T. A. Seagrass dieback in Florida Bay: long-term trends in abundance and growth of turtle grass, Thalassia testudinum. Estuaries 22 , 460–470 (1999).

Harris, R. M. B. et al. Biological responses to the press and pulse of climate trends and extreme events. Nat. Clim. Change 8 , 579–587 (2018).

Jactel, H. et al. Drought effects on damage by forest insects and pathogens: a meta-analysis. Glob. Change Biol. 18 , 267–276 (2012).

Gaylord, M. L. et al. Drought predisposes piñon–juniper woodlands to insect attacks and mortality. N. Phytol. 198 , 567–578 (2013).

Article   CAS   Google Scholar  

McKee, K. L., Mendelssohn, I. A. & Materne, M. D. Acute salt marsh dieback in the Mississippi River deltaic plain: a drought-induced phenomenon? Glob. Ecol. Biogeogr. 13 , 65–73 (2004).

Breshears, D. D. et al. Tree dieback in response to global change-type drought: mortality insights from a decade of plant water potential measurements. Front. Ecol. Environ. 7 , 185–189 (2009).

van der Putten, W. H. Die-back of Phragmites australis in European wetlands: an overview of the European research programme on reed die-back and progression (1993–1994). Aquat. Bot. 59 , 263–275 (1997).

Breshears, D. D. et al. The critical amplifying role of increasing atmospheric moisture demand on tree mortality and associated regional die-off. Front. Plant Sci. 4 , 1–4 (2013).

Matusick, G. et al. Chronic historical drought legacy exacerbates tree mortality and crown dieback during acute heatwave-compounded drought. Environ. Res. Lett. 13 , 095002 (2018).

Davidson, N. C. How much wetland has the world lost? Long-term and recent trends in global wetland area. Mar. Freshw. Res. 65 , 934–941 (2014).

Li, X., Bellerby, R., Craft, C. & Widney, S. E. Coastal wetland loss, consequences, and challenges for restoration. Anthropocene Coasts 1 , 1–5 (2018).

Google Scholar  

Kirwan, M. L. et al. Limits on the adaptability of coastal marshes to rising sea level. Geophys. Res. Lett. 37 , 23 (2010).

Schuerch, M. et al. Future response of global coastal wetlands to sea-level rise. Nature 561 , 231–234 (2018).

Article   ADS   CAS   PubMed   Google Scholar  

Morris, J. T., Sundareshwar, P. V., Nietch, C. T., Kjerfve, B. & Cahoon, D. R. Responses of coastal wetlands to rising sea level. Ecology 83 , 2869–2877 (2002).

Kirwan, M. & Megonigal, J. Tidal wetland stability in the face of human impacts and sea-level rise. Nature 504 , 53–60 (2013).

Cahoon, D. R., McKee, K. L. & Morris, J. T. How plants influence resilience of salt marsh and mangrove wetlands to sea-level rise. Estueries Coasts 44 , 883–898 (2021).

DeLaune, R. D., Baumann, R. H. & Gosselink, J. G. Relationships among vertical accretion, coastal submergence, and erosion in a Louisiana Gulf Coast marsh. J. Sediment. Res. 53 , 147–157 (1983).

Lovelock, C. E. et al. Mangrove dieback during fluctuating sea levels. Sci. Rep. 7 , 1680 (2017).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Silliman, B. R., Van De Koppel, J., Bertness, M. D., Stanton, L. E. & Mendelssohn, I. A. Drought, snails, and large-scale dieback of southern US salt marshes. Science 310 , 1803–1806 (2005).

Gambrell, R. P. & Patrick, W. H. Chemical and microbiological properties of anaerobic soils and sediments. In Plant Life in Anaerobic Environments (eds Hook, D. D. & Crawford, R. M. M.), 375–423 (Ann Arbor Science; Ann Arbor, MI, USA, 1978).

Portnoy, J. & Giblin, A. Effects of historic tidal restrictions on salt marsh sediment chemistry. Biogeochemistry 36 , 275–303 (1997).

Armstrong, J. & Armstrong, W. An overview of the effects of phytotoxins on Phragmites australis in relation to die-back. Aquat. Bot. 69 , 251–268 (2001).

Rea, N. Water levels and Phragmites : decline from lack of regeneration or dieback from shoot death. Folia Geobot. 31 , 85–90 (1996).

Stagg, C. L. et al. Extreme precipitation and flooding contribute to sudden vegetation dieback in a coastal salt marsh. Plants 10 , 1841 (2021).

Article   PubMed   PubMed Central   Google Scholar  

Nechwatal, J., Wielgoss, A. & Mendgen, K. 2008. Flooding events and rising water temperatures increase the significance of the reed pathogen Pythium phragmitis as a contributing factor in the decline of Phragmites australis . Hydrobiologia 613 , 109–115 (2008).

Cronin, J. T., Johnston, J. & Diaz, R. Multiple potential stressors and dieback of Phragmites australis in the Mississippi River Delta, USA: implications for restoration. Wetlands 40 , 2247–2261 (2020).

Knight, I. A. et al. Invasion of Nipponaclerda biwakoensis (Hemiptera: Aclerdidae) and Phragmites australis die-back in southern Louisiana, USA. Biol. Invasions 20 , 2739–2744 (2018).

Coleman, J. M., Roberts, H. H. & Stone, G. W. Mississippi River delta: an overview. J. Coast. Restor. 14 , 698–717 (1998).

Batker, D. et al. The importance of Mississippi Delta restoration on the local and national economies. In Perspectives on the Restoration of the Mississippi Delta: The Once and Future Delta (eds Day, J. W., Kemp, G. P., Freeman, A. M. & Muth, D. P.) 141–153 (Springer, 2014).

Walker, H. J., Coleman, J. M., Roberts, H. H. & Tye, R. S. Wetland loss in Louisiana. Geogr. Ann. 69A , 189–200 (1987).

Bourne, J. Louisiana’s vanishing wetlands: going, going. Science 289 , 1860–1863 (2000).

Article   CAS   PubMed   Google Scholar  

Day, J. W. et al. Pattern and process of land loss in the Mississippi Delta: a spatial and temporal analysis of wetland habitat change. Estuaries 23 , 425–438 (2000).

Blum, M. D. & Roberts, H. H. Drowning of the Mississippi Delta due to insufficient sediment supply and global sea-level rise. Nat. Geosci. 2 , 488–491 (2009).

Article   ADS   CAS   Google Scholar  

Couvillion, B. R., Beck, H., Schoolmaster, D. & Fischer, M. Land Area Change in Coastal Louisiana (1932 to 2016). Map 3381 (U.S. Geological Survey, 2017).

Jankowski, K. L., Törnqvist, T. E. & Fernandes, A. M. Vulnerability of Louisiana’s coastal wetlands to present-day rates of relative sea-level rise. Nat. Commun. 8 , 14792 (2017).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Törnqvist, T. E., Jankowski, K. L., Li, Y. X. & González, J. L. Tipping points of Mississippi Delta marshes due to accelerated sea-level rise. Sci. Adv. 22 , eaaz5512 (2020).

Saintilan, N. et al. Constraints on the adjustment of tidal marshes to accelerating sea level rise. Science 377 , 523–527 (2022).

Olson, K. & Suski, C. Mississippi River Delta: land subsidence and coastal erosion. Open J. Soil Sci. 11 , 139–163 (2021).

DeLaune, R. D., Nyman, J. A. & Patrick Jr, W. H. Peat collapse, ponding and wetland loss in a rapidly submerging coastal marsh. J. Coast. Res. 10 , 1021–1030 (1994).

Day, J. et al. Vegetation death and rapid loss of surface elevation in two contrasting Mississippi delta salt marshes: the role of sedimentation, autocompaction and sea-level rise. Ecol. Eng. 37 , 229–240 (2011).

Temmerman, S., Moonen, P., Schoelynck, J., Govers, G. & Boua, T. J. Impact of vegetation dieback on spatial flow patterns over a tidal marsh. Geophys. Res. Lett. 39 , 3 (2012).

United States Drought Monitor. National Weather Service Climate Prediction Center. National Oceanographic and Atmospheric Administration. (2012) https://www.cpc.ncep.noaa.gov/products/Drought/ .

Rippey, B. R. The U.S. drought of 2012. Weather Clim. Extrem. 10 , 57–64 (2015).

National Hurricane Center, US Department of Commerce, National Oceanic and Atmospheric Administration. (2012) https://www.nhc.noaa.gov/ .

Tilman, D., May, R. M., Lehman, C. L. & Nowak, M. A. Habitat destruction and the extinction debt. Nature 371 , 65–66 (1994).

Gardner, T. A., Côté, I. M., Gill, J. A., Grant, A. & Watkinson, A. R. Hurricanes and Caribbean coral reefs: impacts, recovery patterns, and role in long‐term decline. Ecology 86 , 174–184 (2005).

Galloway, J. N. et al. Nitrogen cycles: past, present, and future. Biogeochemistry 70 , 153–226 (2004).

Khan, F. A. & Ansari, A. A. Eutrophication: an ecological vision. Bot. Rev. 71 , 449–482 (2005).

Day, J. W. et al. Consequences of climate change on the ecogeomorphology of coastal wetlands. Estuar. Coasts 31 , 477–491 (2008).

Gigante, D. et al. New occurrence of reed bed decline in southern Europe: do permanent flooding and chemical parameters play a role? C. R. Biol. 337 , 487–498 (2014).

Article   PubMed   Google Scholar  

Hester, M. W. & Mendelssohn, I. A. Long-term recovery of a Louisiana brackish marsh plant community from oil-spill impact: vegetation response and mitigating effects of marsh surface elevation. Mar. Environ. Res. 49 , 233–254 (2000).

Silliman, B. R. & Bertness, M. D. Shoreline development drives invasion of Phragmites australis and the loss of plant diversity on New England salt marshes. Conserv. Biol. 18 , 1424–1434 (2004).

Li, B. et al. Spartina alterniflora invasions in the Yangtze River estuary, China: an overview of current status and ecosystem effects. Ecol. Eng. 35 , 11–520 (2009).

Gedan, K. B. & Bertness, M. D. How will warming affect the salt marsh foundation species Spartina patens and its ecological role? Oecologia 164 , 79–487 (2010).

Silliman, B. R. et al. Degradation and resilience in Louisiana salt marshes after the BP–Deepwater Horizon oil spill. Proc. Nat. Acad. Sci. USA 109 , 11234–11239 (2012).

Khanna, S. et al. Marsh loss due to cumulative impacts of Hurricane Isaac and the Deepwater Horizon oil spill in Louisiana. Remote Sens. 9 , 169 (2017).

Rabalais, N. N. et al. Nutrient changes in the Mississippi River and system responses on the adjacent continental shelf. Estuaries 19 , 386–407 (1996).

Brix, H. The European research project on reed die-back and progression (EUREED). Limnologica 29 , 5–10 (1999).

Webb, E. C., Mendelssohn, I. A. & Wilsey, B. J. Causes for vegetation dieback in a Louisiana Salt Marsh: a bioassay approach. Aquat. Bot. 51 , 281–289 (1995).

Guo, Z. et al. Extremely low genetic diversity across mangrove taxa reflects past sea level changes and hints at poor future responses. Glob. Change Biol. 24 , 1741–1748 (2018).

Lee, H. et al. Vegetation restoration following dieback of Phragmites australis in the Mississippi River Delta, USA. Wetlands 43 , 98 (2023).

Van Belzen, J. et al. Vegetation recovery in tidal marshes reveals critical slowing down under increased inundation. Nat. Commun. 8 , 15811 (2017).

Borum, J. et al. The potential role of plant oxygen and sulphide dynamics in dieback events of the tropical seagrass, Thalasia testudinum. J. Ecol. 93 , 148–158 (2005).

Armstrong, J., Afreen-Zobayed, F. & Armstrong, W. Phragmites die-back: sulfide and acetic acid induced bud and root death, lignifications, and blockages within the aeration and vascular system. N. Phytol. 134 , 105–113 (1996).

Lissner, J. & Schierup, H.-H. Effect of salinity on the growth of Phragmites australis . Aquat. Bot. 55 , 247–260 (1997).

Achenbach, L., Eller, F., Nguyen, L. X. & Brix, H. Differences in salinity tolerance of genetically distinct Phragmites australis clones. AoB Plants 5 , plt019 (2013).

Wiman, C., Hamilton, B., Dee, S. G. & Muñoz, S. E. Reduced lower Mississippi River discharge during the Medieval era. Geophys. Res, Lett. 48 , e2020GL091182 (2021).

Benoit, L. K. & Askins, R. A. Impact of the spread of Phragmites on the distribution of birds in Connecticut tidal marshes. Wetlands 19 , 194–208 (1999).

Craig, R. J. & Beal, K. G. The influence of habitat variables on marsh bird communities of the Connecticut River estuary. Wilson Bull. 104 , 295–311 (1992).

Elsey, R. M., Trosclair, P. L. & Glenn, T. C. Nest-site fidelity in American alligators in a Louisiana coastal marsh. Southeast. Nat. 7 , 737–743 (2008).

Bart, D. & Hartman, J. M. Environmental determinants of Phragmites australis expansion in a New Jersey salt marsh: an experimental approach. Oikos 89 , 59–69 (2000).

Brix, H., Sorrell, B. K. & Schierup, H.-H. Gas fluxes achieved by in-situ convective flow in Phragmites australis . Aquat. Bot. 54 , 151–163 (1996).

Coleman, D. J. & Kirwan, M. L. The effect of a small vegetation dieback event on salt marsh sediment transport. Earth Surf. Process. Landf. 44 , 944–952 (2019).

Coastal Protection and Restoration Authority (CPRA) of Louisiana. Coastwide reference monitoring system-wetlands monitoring data. Retrieved from Coastal Information Management System (CIMS) database http://cims.coastal.louisiana.gov (2022).

Folse, T. M. et al. A Standard Operating Procedures Manual for the Coastwide Reference Monitoring System-Wetlands: Methods for Site Establishment, Data Collection, and Quality Assurance/Quality Control 226 (Louisiana Coastal Protection and Restoration Authority, Baton Rouge, LA, 2018).

Cretini, K. F., Visser, J. M., Krauss, K. W. & Steyer, G. D. Development and use of a floristic quality index for coastal Louisiana marshes. Environ. Monit. Assess. 184 , 2389–2403 (2012).

White, D. A., Hauber, D. P. & Hood, C. S. Clonal differences in Phragmites australis from a unique wetland landscape—The Mississippi River delta, USA. Southeast. Nat. 3 , 531–544 (2004).

Chabreck, R. H. Vegetation, water and soil characteristics of the Louisiana coastal region (664). Retrieved from https://digitalcommons.lsu.edu/agexp/147 (1972).

Morton, R. A., Bernier, J. C., Barras, J. A. & Ferina, N. F. Historical subsidence and wetland loss in the Mississippi Delta plain. Trans. Gulf Coast Assoc. Geol. Soc. 55 , 555–571 (2005).

Lambertini, C. et al. Tracing the origin of Gulf Coast Phragmites (Poaceae): a story of long-distance dispersal and hybridization. Am. J. Bot. 99 , 538–551 (2012).

Cahoon, D. R. et al. High-precision measurements of wetland sediment elevation: II. The rod surface elevation table. J. Sed. Res. 72 , 734–739 (2002).

Welch, B. L. On the comparison of several mean values: an alternate approach. Biometrika 38 , 330–336 (1951).

Article   MathSciNet   Google Scholar  

Osland, M. J., Enwright, N., Day, R. H. & Doyle, T. W. Winter climate change and coastal wetland foundation species: salt marshes vs. mangrove forests in the southeastern United States. Glob. Change Biol. 19 , 1482–1494 (2013).

Osland, M. J., Enwright, N. & Stagg, C. L. Freshwater availability and coastal wetland foundation species: ecological transitions along a rainfall gradient. Ecology 95 , 2789–2802 (2014).

Download references

Acknowledgements

This work was funded by the U.S. Department of Agriculture Cooperative Agreement (Award number: AP20PPQS&T00C189 to Rodrigo Diaz and Tracy Quirk). Fieldwork was conducted and data was provided through the Coastwide Reference Monitoring System. We are grateful for the support and discussions from J. Andy Nyman (Louisiana State University), Leigh Anne Sharpe (Coastal Protection and Restoration Authority), and Barret Fortier (US Fish and Wildlife Service). The manuscript was greatly improved by the insightful and excellent recommendations of Dr. Matthew Kirwan and one anonymous reviewer.

Author information

Austin Lynn

Present address: College of Agriculture and Life Sciences, Cornell University, Ithaca, NY, 14850, USA

Authors and Affiliations

Department of Oceanography and Coastal Sciences, Louisiana State University, Baton Rouge, LA, USA

Tracy Elsey-Quirk, Austin Lynn, Michael Derek Jacobs, Lixia Wang, Haosheng Huang & Dubravko Justic

Department of Entomology, Louisiana State University Agricultural Center, Baton Rouge, LA, USA

Rodrigo Diaz

Department of Biological Sciences, Louisiana State University, Baton Rouge, LA, USA

James T. Cronin

You can also search for this author in PubMed   Google Scholar

Contributions

T.Q. conceived and organized the study, analyzed the data, and created the figures; R.D. received the funding; L.W. contributed figures; T.Q., A.L., M.D.M., R.D., J.T.C., L.W., H.H., and D.J. wrote the paper.

Corresponding author

Correspondence to Tracy Elsey-Quirk .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks Denise Reed and Matthew Kirwan for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, reporting summary, description of additional supplementary files, peer review file, supplementary data 1, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Elsey-Quirk, T., Lynn, A., Jacobs, M.D. et al. Vegetation dieback in the Mississippi River Delta triggered by acute drought and chronic relative sea-level rise. Nat Commun 15 , 3518 (2024). https://doi.org/10.1038/s41467-024-47828-x

Download citation

Received : 17 August 2023

Accepted : 15 April 2024

Published : 25 April 2024

DOI : https://doi.org/10.1038/s41467-024-47828-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• International edition
• Australia edition
• Europe edition

China floods: four killed in Guangdong sparking concerns over extreme weather defences

Heavy rainstorms in the densely populated Pearl River Delta have left large parts of Guangdong province underwater

Heavy rainstorms that swept across southern China over the weekend killed at least four people as floods swamped cities in the densely populated Pearl River Delta, state media reported.

A search was under way for 10 others missing after record-breaking rains sparked concerns about the region’s defences against bigger deluges induced by extreme weather events.

By Monday, about 110,000 people had been evacuated across the province, while 25,800 people were in emergency shelters, according to Xinhua. In Guangzhou, the capital of Guangdong province, the government said the city had logged a cumulative rainfall of 60.9cm in April, the highest monthly rainfall since record-keeping began in 1959.

The country’s highest-level red rainstorm warning was issued for parts of Guangdong, including the megacity of Shenzhen, the city’s meteorological observatory said. The areas listed were experiencing “heavy to very heavy downpours”, the weather agency said, adding the risk of flash floods was “very high”.

The official Xinhua news agency said three people died in Zhaoqing city while one rescuer died in Shaoguan city. It didn’t provide details about when or how they died. The two cities in Guangdong province are among the worst hit areas of sustained torrential rains that began late last week.

Footage on state broadcaster CCTV showed rescuers in rubber boats evacuating residents from inundated shopping streets and residential areas.

Floods also battered neighbouring Jiangxi province where local media reported 459 people had been evacuated, while rains and floods have affected 1,500 hectares of crops and caused financial losses of more than 41 million yuan ($5.7m).

Guangdong, once dubbed the “factory floor of the world”, is prone to summer floods. Its defences against disruptive floods were severely tested in June 2022 when the province was pounded by the heaviest downpours in six decades. Hundreds of thousands of people were evacuated.

Since Thursday, Guangdong has been battered by unusually heavy , sustained and widespread rainfall, with powerful storms ushering in an earlier-than-normal start to the province’s annual flooding season in May and June.

Over the weekend, waterways in the province overflowed, including in some villages where flood waters reached the second storey of houses after washing out paddy and potato fields.

In Qingyuan, a relatively small city of 4 million, rescuers tackled neck-high waters to extract residents including an elderly lady trapped in waist-deep water in an apartment building.

Others remained on the upper floors of their houses, waiting for the waters to recede as friends delivered food by boat.

Weather events in China have become more intense and unpredictable because of global warming, scientists say, with record-breaking rainfall and drought assailing the world’s second-largest economy, often at the same time.

In Qingyuan residents counted their losses, with one farmer telling Reuters that their rice fields had been “fully flooded”.

“I won’t be making any money this year, I will be making losses,” Huang Jingrong said, estimating his losses at about 100,000 yuan ($13,800).

“What can we do? We won’t get reimbursed for our losses.”

The Associated Press and Reuters contributed to this report

• Extreme weather

More on this story

Kenya floods: tourists evacuated from Maasai Mara after river bursts banks

Parts of Kenya hit by torrential rain and deadly floods – video

‘Water everywhere’: Shropshire farmers race to salvage harvest after record rain

Kenya flood death toll rises as more torrential rain forecast

Weather tracker: heavy rainfall causes flooding and death in east Africa

Global heating and urbanisation to blame for severity of UAE floods, study finds

‘Fields are completely underwater’: UK farmers navigate record rainfall

Millions at risk of floods in China’s Guangdong province after heavy rain

Most viewed.

• Mission, vision and values
• Advisory board
• Anonymous sources policy
• Write for us
• Partner with us
• Make a complaint
• Newsletters
• Clean Energy Frontier

Southern Africa drought flags dilemma for loss and damage fund

Scientists blame the current drought on El Niño – which could exclude those affected from receiving aid for climate-change damage

Residents of Pumula East township walk home after fetching water from a well, (March 7, 2024. REUTERS/KB Mpofu)

Since January, swathes of southern Africa have been suffering from a severe drought, which has destroyed crops, spread disease and caused mass hunger. But its causes have raised tough questions for the new UN fund for climate change losses.

Christopher Dabu, a priest in Lusitu parish in southern Zambia, one of the affected regions, said that because of the drought, his parishioners “have nothing”- including their staple food.

“Almost every day, there’s somebody who comes here to knock on this gate asking for mielie meal, [saying] ‘Father, I am dying of hunger’,” Dabu told Climate Home outside his church last month.

The government and some humanitarian agencies were quick to blame the lack of rain on climate change .

Zambia’s green economy minister Collins Nzovu told reporters in March, “there’s a lot of infrastructure damage as a result of climate change”. He added that the new UN-backed loss and damage fund, now being set up to help climate change victims, “must speak to this”.

Reverend Christopher Dabu outside his church in Lusitu, Zambia (Photo: Joe Lo)

But last week, scientists from the World Weather Attribution (WWA) group published a study which found that “climate change did not emerge as the significant driver” of the current drought affecting Zambia, Zimbabwe, Malawi, Angola, Mozambique and Botswana.

Instead, they concluded that the El Niño phenomenon – which occurs every few years with warming of sea surface temperatures in the eastern Pacific Ocean – was the drought’s “key driver”. They said the damage was worsened by the vulnerabilities of the countries affected, including reliance on rain-fed farming rather than irrigation.

Nonetheless, briefing journalists on the study, co-authors Joyce Kimutai and Friederike Otto said climate change does make El Niños stronger and more frequent – and therefore could be playing an indirect role in the southern African drought. Otto noted that climate change “might have a small role but not a big one”.

While WWA studies have often found that disasters like this are driven by climate change, there have been other cases where they have played down that link – as with droughts in Brazil in 2014 and Madagascar in 2021, and floods in Italy in 2023.

The complex nature of the science raises a dilemma for those now designing the fledgling loss and damage fund.

Its board holds its first meeting in Abu Dhabi this week. In three days of talks, the board’s 26 members will discuss the fund’s name and how to decide where it will be hosted and who will lead it. Trickier issues like the role of climate change attribution will be left to future meetings.

Climate Home spoke to several experts and two of the fund’s board members, whose opinions were divided on whether the link between climate change and a particular disaster should have to be proven before funds are dished out to affected communities.

Droughts and climate change

Egyptian climate negotiator Mohamed Nasr, a member of the new fund’s board, said he thought triggers for funding “would include the climate relation to the losses and damages”.

But to judge that connection, he said the board would “rely on confirmed science per the Intergovernmental Panel on Climate Change (IPCC) and United Nations Environment Programme (UNEP) rather than individual studies”. He said the IPCC and UNEP “provide the scientific reference needed as they bring all views and assess the credibility and scientific basis”.

Peak COP? UN looks to shrink Baku and Belém climate summits

The IPCC does not do original research, including attribution studies, but every five to seven years it compiles existing research to reach conclusions about climate change, including its impacts. The last IPCC report focused on that topic in 2022 said “increases in drought frequency and duration are projected over large parts of southern Africa”.

UNEP currently does not conduct attribution studies, with a spokesperson saying this was “due to resource constraints” but adding “we hope to do more in the future”.

Another loss and damage fund board member, who wanted to remain anonymous, said the fund should only disburse money for loss and damage caused by climate change. But they asserted that due to the “chicken and egg” link between climate change and El Niño, the current southern African drought is climate-driven and so its victims should be entitled to funding.

‘Theoretical disputes’

Mattias Söderberg, who works for humanitarian organisation DanChurchAid – which has been defining and addressing loss and damage since 2019 – said attribution “is not always easy”.

But, he added, “people facing disasters should not be left behind because of theoretical disputes about attribution”.

Speaking ahead of a visit to a Kenyan refugee camp for people displaced by what he calls “loss and damage and climate-related conflicts”, he said, “I’m pretty sure they will be frustrated if they knew funding to help them cope could be questioned.”

The loss and damage fund, with advice from scientists, should draw up categories of disaster that tend to be driven by climate change – like heatwaves and droughts but excluding earthquakes which are not, he added.

Tensions rise over who will contribute to new climate finance goal

Zoha Shawoo, who researches loss and damage at the Stockholm Environment Institute, said that even if climate change played only a small role in the latest southern Africa drought, previous climate disasters had made the region’s people more vulnerable to the drought.

In addition, the current dry spell leaves them more vulnerable to future climate disasters, she added. “If they don’t receive financial support for recovery, future losses and damages will be a lot worse,” she said.

Gernot Laganda, director for climate and resilience at the UN’s World Food Programme, said that a formal attribution requirement for the loss and damage fund feels like “overkill” for a still relatively small fund. Transaction costs should be kept as low as possible, he added.

Kimutai, who worked on the WWA study, said she was confident the group had enough data to reach its conclusions on this particular drought. But she told a webinar hosted by the CGIAR agricultural research centre last month that a lack of data in many poorer countries means a funding requirement of attribution to global warming would be “detrimental to climate justice”.

In 2022, WWA was unable to work out the role of climate change in a drought in the Sahel region of Africa, partly blaming a lack of data. One of the drought-hit countries was Mali – which is three times the size of Germany. Mali has just 13 active weather stations, while Germany has 200, according to Bloomberg .

Limiting frontline voices in the Loss and Damage Fund is a recipe for disaster

Kimutai added that, besides data, there is a lack of expertise in doing these kinds of studies in the Global South.

Any moves to deny funds to vulnerable people impacted by drought – whatever the causes – are likely to be met with anger. Speaking to journalists about the southern Africa emergency a few days after the WWA study was issued, Chikwe Mbweeda, Zambia director for the aid agency CARE, said that “for us, we definitely understand that [the drought] is coming from the climate change effects”.

(Reporting by Joe Lo; editing by Megan Rowling)

Read more on: Loss and damage | Drought | Loss and damage | Science Loss and damage

The climate crisis - and how we confront it - is THE story of our lifetimes. At Climate Home News, we believe journalism can make a difference. Our mission is to produce original reporting that informs, engages and inspires action. This takes time and expertise. Support our work today with a donation or by subscribing to our daily newsletter for exclusive extra content.